Differences in community structure
If the study interest is in community structure and not in richness estimates then preservation choices become even less important. In fact, the similarity of communities obtained through the different preservation conditions is surprisingly high (see Procrustes correlation coefficients; Fig. 4); the amount of variation explained by preservation conditions was much lower than the observed differences among habitats (see redundancy analysis), and multivariate analyses consistently allowed to detect community differences among habitats (Fig. 3). In other words, metabarcoding is able to identify the ecological differences among sites, independently of the preservation approach. Even though metabarcoding analyses are sensitive to methodological choices, estimates of relationships between diversity and the environment are often robust (Calderón‐Sanou et al., 2020; Ji et al., 2013), and this is a very good news if we want to apply these approaches to broad-scale monitoring programs, aiming at assessing the effects of environmental changes.
Bacteria were the only taxon for which we detected significant differences between the “control” and the preservation conditions, with ≈3% of variability explained by differences between the “control” and the desiccation treatments. The observed differences most likely refer to some taxa that are affected by the dry conditions and could lead to a overrepresentation of some taxa that are more resistant under these conditions. Differences between the desiccation treatments and the control were small, and only one out of 660 MOTUs showed a significant variation in abundance with the control. This MOTU (belonging to the Bacteroidetes phylum of bacteria, see Supporting Information) was generally abundant in the control and preservation conditions 2 and 3 (average frequency of reads around 10%) while it drastically decreased under preservation conditions 4 and 5 (Fig. S1).
Nevertheless, the significant differences observed for some taxa and preservation conditions stress the importance of selecting the preservation method before starting a monitoring program and using it consistently through the whole monitoring, to avoid confusion between the effects of methods and of environmental changes.