Influence of preservation methods on richness estimates
Preservation methods generated some small but significant differences in MOTUs richness compared to what is observed in the “control”, with some contrasting effects across taxa. When considering all the MOTUs, none of the preservation conditions yielded estimates of alpha-diversity identical to the “control”. For instance, just six hours at room temperature caused a significant decrease of MOTUs richness in fungi. It has been shown that estimates of alpha-diversity using metabarcoding are extremely sensitive to methodological choices (Calderón‐Sanou et al., 2020). Our study underlines that even preservation for a very short time can further affect the detection of rare MOTUs and highlights the sensitivity of fungi to preservation at room temperature (Delavaux, Bever, Karppinen, & Bainard, 2020). MOTUs richness of all the taxa was also affected by preservation at 4°C, which caused a slight increase of MOTUs richness for bacteria and eukaryotes, and a slight decrease for fungi. The effect of temperature and time storage in fungal and bacterial growth has already been proved (see e.g. Orchard et al., 2017; Pettersson & Bååth, 2003). Despite this, in addition to temperature, we can expect that other parameters such as initial soil moisture and pH influences bacterial growth (Bååth & Arnebrant, 1994; Drenovsky, Vo, Graham, & Scow, 2004; Fernández-Calviño & Bååth, 2010; Kaiser et al., 2016) with a combined effect. Finally, drought affects the richness of microbial communities in soil ecosystems with differential effects across taxa depending on their ecology (Evans, Wallenstein, & Burke, 2014; Meisner, Jacquiod, Snoek, Ten Hooven, & van der Putten, 2018; Ochoa-Hueso et al., 2018), and three weeks of preservation with silica gel generally reduced the observed MOTUs richness in bacteria and fungi, while it increased the richness of eukaryotes.
However, our study also shows that specific caution is mostly necessary when rare MOTUs are of interest. The exclusion of rare MOTUs strongly reduced differences between optimal conditions and different preservation. Remaining effects were much weaker for bacteria and fungi, and disappeared for eukaryotes (Fig. 2). This suggests that the effect of preservation approach on taxonomic richness mostly occur on rare species, as already suggested by Meisner et al. (2018) for microbial communities. Several authors have shown that eDNA metabarcoding does not represent the best tool for the detection of rare MOTUs, as some rare MOTUs remain undetected, while many sequences detected at rare frequency are artifactual (Brown et al., 2015). Estimates of α-diversity should therefore always taken with caution, and indices that underweight rare MOTUs (e.g. Shannon or Simpson diversity) can provide more robust estimates (Brown et al., 2015; Calderón‐Sanou et al., 2020; Bálint et al., 2016).