The effects of spatial and temporal replicate sampling on eDNA metabarcoding

Abstract

Background The heterogeneous nature of environmental DNA (eDNA) and its effects on species detection and community composition estimates has been highlighted in several studies in the past decades. Mostly in the context of spatial distribution over large areas, in fewer occasions looking at spatial distribution within a single body of water. Temporal variation of eDNA, similarly, has mostly been studied as seasonality, observing changes over large periods of time, and often only for small groups of organisms such as fish and amphibians. Methods We analyzed and compared small-scale spatial and temporal variation by sampling eDNA from two small, isolated dune lakes for 20 consecutive weeks. Metabarcoding was performed on the samples using generic COI primers. Molecular operational taxonomic unit (MOTUs) were used to assess dissimilarities between spatial and temporal replicates. Results Our results show large differences between samples taken within one lake at one point in time, but also expose the large differences between temporal replicates, even those taken only 1 week apart. Furthermore, between-site dissimilarities showed a linear correlation with time frame, indicating that between-site differences will be inflated when samples are taken over a period of time. We also assessed the effects of PCR replicates and processing strategies on general patterns of dissimilarity between samples. While more inclusive PCR replicate strategies lead to higher richness estimations, dissimilarity patterns between samples did not significantly change. Conclusions We conclude that the dissimilarity of temporal replicates at a 1 week interval is comparable to that of spatial replicate samples. It increases, however, for larger time intervals, which suggests that population turnover effects can be stronger than community heterogeneity. Spatial replicates alone may not be enough for optimal recovery of taxonomic diversity, and cross-comparisons of different locations are susceptible to inflated dissimilarities when performed over larger time intervals. Many of the observed MOTUs could be classified as either phyto- or zooplankton, two groups that have gained traction in recent years as potential novel bio-indicator species. Our results, however, indicate that these groups might be susceptible to large community shifts in relatively short periods of time, highlighting the need to take temporal variations into consideration when assessing their usability as water quality indicators.

Assessing the impact of the threatened crucian carp (Carassius carassius) on pond invertebrate diversity - a comparison of conventional and molecular tools

“…DNA metabarcoding of aquatic invertebrate samples has proven relatively successful, with applications in biomonitoring (Andújar et al, 2017;Elbrecht, Vamos, Meissner, Aroviita, & Leese, 2017a;Emilson et al, 2017;Lobo, Shokralla, Costa, Hajibabaei, & Costa, 2017). Use of eDNA metabarcoding for invertebrate assessment in freshwater rivers (Blackman et al, 2017;Carew, Kellar, Pettigrove, & Hoffmann, 2018;Deiner, Fronhofer, Mächler, Walser, & Altermatt, 2016;Klymus, Marshall, & Stepien, 2017;Leese et al, 2020), streams (Macher et al, 2018), and lakes (Klymus et al, 2017) is also gaining traction, but there are currently few published studies that have used metabarcoding for small lake or pond invertebrates (Beentjes, Speksnijder, Schilthuizen, Hoogeveen, & van der Hoorn, 2019).…”

A practical guide to DNA metabarcoding for entomological ecologists

“…For better estimation of biodiversity using a filtration method, three to nine sampling replicates ( c . 300 ml water per replicate) within each study site are recommended (Grey et al ., ; Beentjes et al ., ). Using ethanol precipitation is suitable for small volume water samples (e.g.…”

“…pore size). These constraints may be responsible for observed heterogeneity of biodiversity among samples (Grey et al ., ; Beentjes et al ., ), and therefore a large number of replicate samples may be required for accurate species detection (Ficetola et al ., ). For better estimation of biodiversity using a filtration method, three to nine sampling replicates ( c .…”

Metabarcoding reveals different zooplankton communities in northern and southern areas of the North Sea

“…Processing of reads was conducted using the Galaxy platform (Afgan et al, 2018) following the principal steps of (Beentjes et al, 2019). Samples taken from different depths of the same sampling station were combined to allow for studying the zooplankton community of the entire watercolumn.…”