eDNA metabarcoding: a realistic perspective
Summary
The classic DNA barcoding approach, first described by Hebert et al. in 2003, was not able to fulfill all
the promises made at the time. Some challenges remained; for instance the ability of DNA barcoding to
identify specimens accurately up to species level. Now, the research field of DNA
barcoding has diversified itself and new DNA barcoding methods have been developed. These methods
were not only applicable to taxonomic research, but scientists realized that new forms of DNA barcoding
(eDNA barcoding, metabarcoding and eDNA metbabarcoding) could be very well used in ecological
research. For this thesis, the main promises of (e)DNA (meta)barcoding and the remaining limitations
have been investigated in order to find a realistic perspective on the (e)DNA (meta)barcoding approach.
eDNA could be used for the identification of species when only part of the species is present in the
environment. Another promise of eDNA barcoding is that with eDNA barcoding it should be possible to
identify species which cannot be morphologically identified. eDNA usually involves shorter, degraded
DNA samples and this is highly compatible to a new, advanced method for DNA sequencing: Next-
Generation Sequencing. With NGS, multiple DNA samples could be analyzed simultaneously, fast and
accurately. The development of NGS facilitated the emergence of (e)DNA metabarcoding, which could
potentially identify all the organisms present in an ecosystem. For these new forms of DNA barcoding,
new techniques had to be developed. Ordinary barcodes were too long and minibarcodes were needed;
the use of NGS involves shorter read lengths than Sanger sequencing. Since the emergence of NGS
techniques around 2005, DNA sequencing becomes faster and less expensive and the amount of reads
per run is increasing. All this data needs to be analyzed and bioinformatics tools could help with this.
Although (e)DNA (meta)barcoding seems very promising for the future, there are still many
limitations and gaps that need to be filled. Some limitations concern sampling issues: how large should
the sample be in order to represent the biodiversity in the area correctly? Furthermore, the (e)DNA
needs to be extracted from the soil and this could influence the results. Other limitations concern the
use of different barcodes and reference databases. There is an ongoing search for new primers and
barcodes, and new reference libraries have to be constructed. Techniques used for (e)DNA
(meta)barcoding could also be limiting for achieving good results. PCR and NGS could influence the
relative quantities of (e)DNA and the ratio DNA:species. Bioinformatics tools have to be improved to
process the large amounts of data generated by NGS and for species assignment. Lastly, there are some
extra limitations when (e)DNA is used. Not all species excrete eDNA in their environment, and when the
eDNA is too much degraded it will bias the results. Furthermore, it is hard to be sure whether the eDNA
found in the samples belongs to species which are still present in the area or if it belongs to species
which are already extinct. Due to the limitations mentioned above, it is not yet possible to conduct a reliable eDNA
metabarcoding analysis. However, research is going full steam ahead and the current limitations do not
seem insuperable. Combined with the fact that every year, NGS is becoming faster and cheaper, the
eDNA metabarcoding approach could become available to ecologists around the world.