In a just-released article in the journal Methods in Ecology and Evolution, we investigate the potential of DNA metabarcoding of freshwater macroinvertebrates for the assessment of stream health. With the “DNA metabarcoding” method hundreds of specimen in a sample can be identified, often up to species level. The assessment results of the morphology and DNA based methods were very similar with both methods, while DNA metabarcoding detected more taxa on species level. However, metabarcoding has also shortcomings, for example, the method failed to detect ~ 30% of the taxa identified by morphology. This work is an important step into validation and improvement of the metabarcoding method, and it highlights it’s potential for more rapid and accurate routine monitoring of freshwater streams.
Picture 1: Vasco Elbrecht, Edith Vamos and Florian Leese (from left to right), in cooperation with Kristian Meissner and Jukka Aroviita from the Finnish Environment institute, analysed 18 monitoring samples from Finnish streams.
Additionally, our R package PrimerMiner was featured on the cover page of the journalMethods in Ecology and Evolution.
Picture 2: The cover picture in Methods in Ecology and Evolution shows a macroinverterte sample which is homogenised using liquid nitrogen and a mortar, in order to extract the DNA from the sample.
We are currently focusing on development and optimisation of DNA metabarcoding primers. As part of this process we have developed the R package “PrimerMiner” (GitHub) which can be used to obtain and process sequence data from BOLD and NCBI databases.
Last year we used a freshwater invertebrate mock community to test the standard COI barcoding Folmer primers in a metabarcoding setting and demonstrated that primer bias makes estimation of biomass impossible. Together with our colleagues from France we now tested the same mock samples with a ribosomal marker, which is supposed to show less primer bias due to more conserved primer binding sites. As we show in the new article which was just published in PeerJ, the 16S marker indeed leads to more consistent amplification results than the Folmer primer set when used in DNA metabarcoding.
However, primer bias is still substantial (two orders of magnitude) and thus estimation of biomass still impossible. Furthermore, we are currently lacking comprehensive 16S reference databases for animals, which are available for the COI barcoding marker. First tests show that the COI marker performance can be greatly improved by using metabarcoding optimised primers (we will publish a PrePrint soon). Thus, while the 16S marker might have potential, we focus on using the COI maker for DNA metabarcoding of insects but just not with Folmer primers.
Figure: The tested 16S marker shows less variation in read numbers than the COI Folmer maker, thus allows for better detection of freshwater taxa. However, with improved COI primers better detection rates are possible as well, as first tests demonstrated.
The first step for DNA metabarcoding of invertebrate kick samples from stream ecosystems is sample homogenisation. The samples often contain hundreds of organisms which have to be dried and then homogenised, to assure that the DNA of all specimens in mixed. Only then every individual in the sample can be detected, extracting just a small subset of the ground tissue. We are using the IKA ULTRA-TURRAX Tube Drive control system for tissue grinding, as discussed in a previous video. In this new video we are detailing the complete process from sample handling, drying over night and grinding using invertebrate kick samples from Finland.
Together with our colleges from France we just have published our first PrePrint, where we test the potential of the 16S marker for DNA metabarcoding. We were able to show that freshwater insects are more consistently amplified with the developed 16S primer than with standard Folmer COI barcoding primer. While the taxonomic resolution of the 16S marker remains unknown for insects, the ribosomal marker might be a viable alternative metabarcoding marker when local reference databases can be easily established. For now we will however continue using the COI marker, due to availability of good reference databases and optimised freshwater COI markers which show similar performance.
In January, Vasco Elbrecht was invited to present our latest DNA metabarcoding results at the Museum Koenig (Bonn) and the University of Landau. The talk held in Landau is available on YouTube.
The presented projects will be summited for publication shortly, but we are happy to share primers and information per E-Mail.
Photo: Vasco Elbrecht in front of the Savanna landscape at the museum Koenig.
Last week Kristian Meissner (PhD) from the Finnish Environment Institute, SYKE, visited us to discuss ongoing and future projects. Currently, we process finnish macrozoobenthos samples for metabarcoding. Using our state-of-the-art metabarcoding protocols we are working on a developing more reliable and cheaper water quality assessment routines than possible with morphology-based methods. We would like to thank Kristian for his visit and are looking forward to future cooperations.
Picture 1: The samples from Finland contain hundreds of specimens. For DNA extraction we dry them over night to remove the ethanol in wich the samples were stored.
Florian Leese and Vasco Elbrecht held a 3-day DNA metabarcoding workshop at the University of Duisburg-Essen last week. We explored and explained the challenges and approaches to utilize DNA metabarcoding to answer scientific questions and improve environmental monitoring. After one day of explaining the theoretical background behind our laboratory routines and bioinformatics pipelines, the participants processed a complete high throughput dataset on their own computers in the two-day hands-on bioinformatics lab.
We would like to thank all colleagues and friends attending, for their great discussions and fun time. See you on the next workshop! Maybe end of 2016?
Foto 1: Colleagues and friends attending our first metabarcoding workshop 2015
Foto 2: Vasco explains how tagswitching between samples can cause problems in DNA metabarcoding.
Anna Kästel from the University Koblenz-Landau and Vasco Elbrecht prepared 20 chironomid samples for DNA metabarcoding over the last two weeks. Together with Kathrin Theissinger, Carsten Brühl and Florian Leese we are investigating the effects of Bacillus thuringiensis israelensis (BTI) on chironomids. BTI is used in the Oberrhein area in Germany to control mosquito populations, but effects on non target species are not well studied. With modern DNA matabarcoding methods we are investigating the response of chironomids to BTI.
Picture: Vasco Elbrecht and Anna Kästel with the finished DNA metabarcoding library.