SPM sampling reflection
I thought this week’s discussion on the Diaz et al. suspended particulate matter (SPM) paper was quite exciting. Sediment sampling, including this SPM sediment trap method, is new to me. I deal primarily with water samples, and my target organisms are phytoplankton and bacteria rather than large metazoans. Utilizing SPM, rather than sediment cores, excludes benthic organisms and older sequences (sequences older than when the study period began). It captures sedimentary particles before they become a part of the entire sediment community. These particles seem quite difficult to sample and it seems to me that the authors did a good job developing the collection method. I think that eDNA monitoring via SPM traps is a novel idea with a lot of potential; however, the study design needed refinement. The main issue I had with this study was the way the authors pooled their samples. SPM was collected from the traps each month but were subsequently pooled in to a yearly sample and then subsampled.
I don’t discount the value of a yearly sampling study, and I can see why a yearly eDNA study would sample monthly (to minimize sequence degradation); but for invasive species detection, I would consider it far more useful to detect presence/absence on a smaller timescale than a year. The current study design only tells us if an organism has passed through the region in the last year. This timescale does not seem appropriate for eDNA sampling to me, as it masks seasonal effects and occludes the timing of which organisms appeared in which month. Lumping in this way assumes all months are the same, or that this difference is irrelevant. If this method was used to monitor invasive species, or other time-sensitive taxa, a monthly or shorter sampling regimen would be more informative when biomonitoring. I think Diaz et al. could have kept monthly collections as monthly collections.
One of the issues the authors mention is collecting an adequate amount of eDNA. Were they unable to get an adequate amount from a single month’s trap collection, or was it necessary to pool samples for an extended time? Was this tested in preliminary experiments? I’d like to see some more experimentation with the optimal amount of time a trap can be left out. Something I’ve been wondering about eDNA in general, including this SPM paper: how do we figure out where the eDNA is being transported from before it settles on the trap? Or, how do we source the origin of eDNA? Does this even matter, as long as we know target taxa are in the general area? Additionally, I wonder how varying sediment trap depth affects detection? Theoretically, it would limit particle collections only to organisms above the level of the trap. I wonder if this is true in situ.