Changes in archaeal, bacterial and eukaryal assemblages along a salinity gradient by comparison of genetic fingerprinting methods in a multipond solar saltern

Microbial communities inhabiting a multipond solar saltern were analysed and compared using SSU rRNA polymerase chain reaction (PCR)-based fingerprintings carried out in parallel by four laboratories. A salinity gradient from seawater (3.7%) to NaCl precipitation (37%) was studied for Bacteria, Archaea and Eukarya, and laboratories applied their own techniques and protocols on the same set of samples. Members of all three domains were retrieved from all salt concentrations. Three fingerprinting techniques were used: denaturing gradient gel electrophoresis (DGGE), ribosomal internal spacer analysis (RISA), and terminal-restriction fragments length polymorphism (T-RFLP). In addition, each laboratory used its own biomass collection method and DNA extraction protocols. Prokaryotes were addressed using DGGE and RISA with different 'domain-specific' primers sets. Eukaryotes were analysed by one laboratory using DGGE and T-RFLP, but targeting the same 18S rDNA site. Fingerprints were compared through cluster analysis and non-metric multidimensional scaling plots. This exercise allowed fast comparison of microbial assemblages and determined to what extent the picture provided by each laboratory was similar to those of others. Formation of two main, salinity-based groups of samples in prokaryotes (4-15% and 22-37% salinity) was consistent for all the laboratories. When other clusters appeared, this was a result of the particular technique and the protocol used in each case, but more affected by the primers set used. Eukaryotic microorganisms changed more from pond to pond; 4-5% and 8-37% salinity were but the two main groups detected. Archaea showed the lowest number of bands whereas Eukarya showed the highest number of operational taxonomic units (OTUs) in the initial ponds. Artefacts appeared in the DGGE from ponds with extremely low microbial richness. On the other hand, different 16S rDNA fragments with the same restriction or internal transcribed spacer (ITS) length were the main limitations for T-RFLP and RISA analyses, respectively, in ponds with the highest OTUs richness. However, although the particular taxonomic composition could vary among protocols, the general structure of the microbial assemblages was maintained.