Sample metrics and analyses
Read metrics (counts and number of zOTUs) were acquired using different summarizing functions during bioinformatic processing (grep, obigrep, gawk), and selection functions (tidyverse, Wickham et al., 2019) during taxonomic filtration in R. To isolate the effect of differential sequencing, we generated an additional dataset (referred to as ‘pilot full’) by sub-setting – from the full dataset – the same real samples that were sequenced in the pilot (n = 75). To test how well the different datasets represent prey diversity, we performed rarefaction analyses with the rarecurve function (step = 10) of vegan (Oksanen et al., 2019), and calculated the average number of reads required to discover new zOTUs (rareslope, sample = sample total reads).