4.4 Influence of water temperature and discharge on species and
OTU richness
We neither detected an influence of temperature on overall species nor
on OTU richness, thus indicating that other factors like sampling time
or season had a greater effect on eDNA species and OTU recovery. In
contrast, we found a negative effect of increasing discharge on
hololimnic taxa, Annelida, and Ephemeroptera species richness. For
Ephemeroptera with a p-value close to 0.05 this should be interpreted as
a trend and future studies are needed to verify the effect of discharge.
For hololimnic species, where Annelida were the most species-rich group,
a possible explanation could be that many Annelida specimens being swept
away with high discharge is low as most annelids are embedded within the
sediment and are potentially moving even deeper into the sediment with
high discharge. This could lead to a higher dilution effect and
therefore lower annelid DNA concentrations as specimens are not washed
away.
In the summer months, an extremely low species and OTU richness was
detected for all groups with August 2018 being the month with the lowest
richness. The summer 2018 was outstandingly warm with low rainfall and
high sun exposure. The exceptionally warm temperatures coupled with low
rainfall probably affected abundance and growth rates of the more
sensitive species and therefore causing a decrease in detection rates
but also increased DNAse activity or microbial break DNA down leading to
faster DNA degradation rates could be an influencing factor (Barnes &
Turner, 2016; Nielsen et al., 2007).
Conclusion Our data show how eDNA-based time-series data on a stream
macroinvertebrate community can support ecological research and
biomonitoring of rivers, as we detected seasonal differences and small
temporal shifts in community composition and key functional feeding
groups. As a clear advantage of eDNA metabrcoding, our study detected
many species of often morphologically neglected taxa such as annelids
and chironomids that are often underrepresented in other studies. By
including such taxa in the analysis, eDNA metabarcoding could function
as a complementary tool for long-term monitoring when assessing
differences in macroinvertebrate community composition on a highly
resolved temporal scale. This could help to disentangle variations in
community composition due to recent and extreme changes from gradual
changes, as seen with our ‘seasonal clock’ data which showed gradual
shifts in temporal beta-diversity, and help deriving guidelines for
future eDNA-based monitoring of rivers and streams.