Jiang and colleagues described the microbial diversity of Lake Chaka, a hypersaline lake in China (2). Numerous typographical errors occur in the names of taxa (e.g., “Salinibacter rubber,” “Holomonas,” “Euryarchaeaota,” “Haloanaerobium lacurosei,” and “Halosimplex carlsbadensis”). The “culture-dependent” methods used to assess microbial diversity were inadequate, as only liquid enrichment cultures were used, which would be expected to recover highly biased, relatively homogeneous populations. Direct plating with extended incubation periods has been shown to recover significant fractions of the dominant microbial groups in soil (4) and in salt water (1). Indeed, the success of such methods in cultivating salt lake microbes is eminently demonstrated by the Maras salterns study (3) that was published in the very same issue as that of Jiang et al. The Lake Chaka study gives no estimate of the ratio of haloarchaea to bacteria in the water column, but it is likely that a large proportion of the >106 cells/ml are haloarchaea. However, only three such isolates are described, a number that is insufficient to assess the microbial diversity of the lake and speaks of serious problems with the culture-dependent aspect of the study. The salt concentration in the lake water is stated as 32.5% (i.e., 32.5 g/100 ml), yet ion concentrations given in the lake water chemistry section (Results) add up to about 3%. Even more confusing is that many ions are stated to have two levels, e.g., 4 and 957 mg/liter for K+. Presumably, these are more typographical errors but in this type of study it is important to present the water chemistry correctly. The soluble salt concentration in the water-sediment interface sample (Fig. 2B, 0 cm) was also surprisingly low (∼3%) given that it is in direct contact with the 32.5% salt lake water and the sediment itself is stated to be rich in halite (solid salt).
- Copyright © 2006 American Society for Microbiology
REFERENCES
Authors' Reply
We sincerely apologize for the typographical errors in our earlier paper (2) and have made corrections in an erratum. After the corrections, the total salt concentration comes to 325 g/liter.
A previous study (3) established a stratigraphic sequence in Lake Chaka as a result of progressive evaporation from freshwater to salt-saturated water over the past 50,000 years: freshwater silty sand at the bottom, gypsum (CaSO4 · 2H2O) and mirabilite (Na2SO4 · 10H2O) in the middle, and halite (NaCl) at the top. So depending on the location within the lake, low-salinity sediments can occur either at depth (∼9 m) or at the water-sediment interface. Our sampling location was at the edge, so the low soluble salt concentration of ∼3% is not surprising.
We acknowledge that direct plating with extended incubation periods can recover significant fractions of the dominant microbial groups in certain environmental samples. It is also well established that the use of liquid media to perform culture-dependent studies is not sufficient to study the microbial diversity in environmental samples. In fact, only 0.1 to 10% of the total population present in environmental samples could be cultivated (4) and different media have certain selective characteristics for different groups of organisms (1). It was not our intention to study the microbial diversity based on culture-dependent results. Our conclusions about the microbial diversity in Lake Chaka were based on a culture-independent method (i.e., 16S rRNA gene analysis). So the bias produced by liquid enrichment in our experiments did not at all affect our conclusions about the microbial diversity in Lake Chaka. The reason for using liquid media was to enrich certain groups of halophiles for future studies.
We did not intend to estimate the ratio of haloarchaea to bacteria in the water column. It was not realistic to even guess such a ratio without further work, such as quantitative PCR or fluorescence in situ hybridization. While we apologize for several typographical errors in the paper, we strongly believe that our experimental design is sound and our results are highly significant.
