Estimating required sample sizes for gut content metabarcoding studies of dietary diversity in spiders

Spiders are dominant predators in terrestrial ecosystems, regulating
invertebrate biomass through predation. However, analysing predation
patterns in the field is challenging, especially as spiders are
extra-intestinal liquid feeders and often active at night. Gut content
metabarcoding (GCM) provides a cost-effective method to study the prey
composition of spiders. In this study, we determine the minimum sample
size of spider specimens required to obtain representative estimates of
local prey composition. Six common spider species were studied using GCM
and we show that the number of analysed individuals per spider species
influenced the saturation levels for prey at different taxonomic levels.
To achieve 90% saturation at the prey species level, between 50 and over
150 individuals per spider species had to be analysed, which is
challenging for ecological multi-site studies. Hence we argue for
consideration of coarser taxonomic levels, such as family or order, which
often provide sufficient ecological information to understand
predator-prey interactions. At these levels, saturation can be achieved
with much smaller sample sizes of 4 to 8 individuals per species, making
ecological GCM studies more cost-effective. These results highlight the
benefits of considering the taxonomic resolution in metabarcoding studies
to address ecological research questions. We further provide information
that will facilitate future investigations of predator-prey dynamics, not
just in spiders, but also in other extra-intestinal liquid feeders like
for example ground beetles.