Archaeal and bacterial tetraether lipids in tropical ponds with contrasting salinity (Guadeloupe, French West Indies): Implications for tetraether-based environmental proxies

The occurrence and distribution of archaeal and bacterial glycerol dialkyl glycerol tetraether lipids (GDGTs) in continental saline environments have been rarely investigated. Here, the abundance and distribution of archaeal isoprenoid GDGTs (iGDGTs) and archaeol, and of bacterial branched GDGTs (brGDGTs) in four tropical water ponds of contrasting salinity in two islands from the French West Indies, Grande-Terre and La Desirade, have been determined. The sedimentary distribution of the GDGTs strongly differed between the two islands. Caldarchaeol was largely predominant among iGDGTs in the (hyper)saline ponds from Grande-Terre, suggesting a substantial contribution of iGDGTs derived from methanogenic Archaea. In contrast, both caldarchaeol and crenarchaeol were present in high relative abundance in the low salinity ponds from La Desirade, suggesting that iGDGTs were derived from mixed archaeal communities. In addition, the relative proportion of the most methylated brGDGTs was much higher in Grande-Terre ponds than in La Desirade ponds. The applicability of different proxies based on GDGTs and archaeol was tested for these specific environments. The relative abundance of archaeol vs. caldarchaeol (ACE index) was comparable for the four ponds, independent of salinity, showing that the ACE index might not necessarily track salinity change. Moreover, the relative proportion of caldarchaeol vs. total iGDGTs was unexpectedly observed to increase with salinity, suggesting production of this compound by halophilic Archaea. The supposed high abundance of methanogenic Archaea in Grande-Terre ponds prevented the application of TEX86 as a temperature proxy, whereas TEX86 could be successfully used for local temperature reconstruction in La Desirade ponds. BrGDGTs seem to be produced predominantly in situ (water column and/or sediment) in hypersaline ponds from Grande-Terre, but in La Desirade ponds likely result from a mixture of soil and aquatic sources. In Grande-Terre ponds, brGDGT-derived temperature estimates generated using either soil or lacustrine calibrations were much lower than expected. The mismatch between expected and estimated temperature might be explained by the presence of halophilic microbial communities producing specific brGDGT distributions in the saline ponds from Grande-Terre. The study shows that the sources of brGDGTs, iGDGTs and archaeol (i) may strongly differ in aquatic environments of varying salinity, even at a regional scale, and (ii) have to be constrained before tetraether-derived proxies in such settings can be confidently applied. (C) 2015 Elsevier Ltd. All rights reserved.