A quest for the biological sources of long chain alkyl diols in the western tropical North Atlantic Ocean

Long chain alkyl diols (LCDs) are widespread in the marine water column and sediments, but their biological sources are mostly unknown. Here we combine lipid analyses with 18S rRNA gene amplicon sequencing on suspended particulate matter (SPM) collected in the photic zone of the western tropical North Atlantic Ocean at 24 stations to infer relationships between LCDs and potential LCD producers. The C30 1,15-diol was detected in all SPM samples and accounted for >95 % of the total LCDs, while minor proportions of C28 and C30 1,13-diols, C28 and C30 1,14-diols, as well as C32 1,15-diol were found. The concentration of the C30 and C32 diols was higher in the mixed layer of the water column compared to the deep chlorophyll maximum (DCM), whereas concentrations of C28 diols were comparable. Sequencing analyses revealed extremely low contributions (≈0.1 % of the 18S rRNA gene reads) of known LCD producers, but the contributions from two taxonomic classes with which known producers are affiliated, i.e. Dictyochophyceae and Chrysophyceae, followed a trend similar to that of the concentrations of C30 and C32 diols. Statistical analyses indicated that the abundance of 4 operational taxonomic units (OTUs) of the Chrysophyceae and Dictyochophyceae, along with 23 OTUs falling into other phylogenetic groups, were weakly (r≤0.6) but significantly (p value <0.01) correlated with C30 diol concentrations. It is not clear whether some of these OTUs might indeed correspond to C28−32 diol producers or whether these correlations are just indirect and the occurrence of C30 diols and specific OTUs in the same samples might be driven by other environmental conditions. Moreover, primer mismatches were unlikely, but cannot be excluded, and the variable number of rRNA gene copies within eukaryotes might have affected the analyses leading to LCD producers being undetected or undersampled. Furthermore, based on the average LCD content measured in cultivated LCD-producing algae, the detected concentrations of LCDs in SPM are too high to be explained by the abundances of the suspected LCD-producing OTUs. This is likely explained by the slower degradation of LCDs compared to DNA in the oxic water column and suggests that some of the LCDs found here were likely to be associated with suspended debris, while the DNA from the related LCD producers had been already fully degraded. This suggests that care should be taken in constraining biological sources of relatively stable biomarker lipids by quantitative comparisons of DNA and lipid abundances.