Sequencing; how deep is deep enough?
The brute force concept was initially tested by Piñol et al., (2014) to determine the prey of the Oedothorax fuscus spider. Using a single pooled sample of 109 individual extracts they acquired what they called ‘ample’ prey reads. Most sequence reads originated from the consumer, although the percentage of O. fuscus reads to the total was not reported. Their overall recovery was 61 000 prey reads, translating to 6% of all assigned reads. In our study the prey recovery rates were much lower, with prey sequences accounting for only 0.44% and 1.22% of the assigned reads in the “pilot” and “full” sequencing runs, respectively. Thus, deeper sequencing was needed to acquire enough reads to cover the full diversity of prey species.
Nonetheless, if comparing prey read metrics with recent studies using PNA or other blocking primers (Table 2), it becomes clear that the applied brute force approach yielded a high absolute prey read output (1.2 and 52.2 mill. prey reads), regardless of high consumer losses. Reaching similar numbers with the use of blocking primers, a recent study of Baltic Sea copepods reported 1.2 million prey reads from an unknown number of samples (Zamora-Terol et al., 2020). The use of a PNA blocker also did not eliminate the consumer DNA problem completely, given that the total read count before discarding Maxillopoda ASVs was 12.8 million reads. Hence, despite using blocking primers, 89% of the sequences were uninformative and discarded. Cleary et al., (2017) also acquired a large dataset with a total of 11.3 million reads from 80 samples of Calanus glacialis copepods. Although a PNA probe also here was used to block amplification of Calanus spp., only a smaller portion of the dataset belonged to prey taxa (638 231 reads, 7 975 prey reads per sample). Whether the discarded reads were assigned to contaminant, symbiont or consumer sources was not reported, but we find it likely that the majority was sourced from the consumer DNA. Likewise, a study of Pseudocalanus spp. copepods found 28 000 prey sequences in 46 samples (Cleary et al., 2016). PNA-PCR was used also here to block amplification of the consumer, but as the authors did not report the initial read counts, the success rate of blocking is not known.