Skip to content


DNA based methods for detecting species from various environments have advanced quickly recently. Our EBAI project has been actively involved in this development. However, molecular monitoring methods have been employed mainly in detecting endangered or invasive species, and DNA metabarcoding of macroinvertebrate communities has focused on small-scale comparisons within single systems or countries. Therefore, researchers from the Nordic countries developed a project, SCANDNAnet, to allow testing DNA metabarcoding of macroinvertebrate communities on a large geographic scale and across countries.

From the SCANDNAnet webpage: «SCANDNAnet covers a geographically very large extent by using samples from the annual national monitoring programs of all Nordic countries. The novel advances made during this project can directly be put into use in the national monitoring programs of the Nordic countries and will have far reaching impact in Europe and beyond. Through intensive dialogue with relevant national and international stakeholders our results will help facilitate cost-effective, standardized DNA-based biomonitoring and create a significant societal impact by promoting reliable future aquatic ecosystem status and service management»

In total, SCANDNAnet has 305 samples, of which 50 samples are from Sweden, 48 from Finland, 139 from Norway, 8 from Iceland, and 60 samples are from Denmark. Currently, these samples are dried and homogenized at the NTNU University Museum. After homogenizing, DNA will be extracted from the samples, and a fragment of the barcoding gene COI will be amplified, using primers BF2 and BR2 (Elbrecht and Leese 2017). The samples will be sequenced, using Illumina HiSeq system. This study scales up the study of Elbrecht et al. (2017) that showed very promising results, using Finnish stream monitoring samples.


Participating institutes:


Finnish Environment Institute SYKE - SCANDNAnet, Finland

Swedish University of Agricultural Sciences, Sweden

NTNU University Museum, Department of Natural History, Norway

Norwegian Institute for Nature Research (NINA), Norway

Norwegian Institute for Water Research (NIVA), Norway

Marine and Freshwater Research Institute, Iceland

Aarhus University, Denmark

University of Duisburg Essen, Germany

University of Guelph, Canada

In sparsely populated countries, many freshwater systems are fairly remote and reachable only by, for example, walking (Fig.1). Therefore, acquiring information about the biological status of these remote systems may pose a challenge if sampling requires carrying a lot of sampling equipment to the site and a lot of sample material back to the lab. This has been the case for traditional monitoring of freshwater invertebrate communities, which is based on extensive sampling of the bottom fauna and storing the collected specimens in ethanol.

Fig. 1. Torbjørn and Markus walking to a sampling site in the Rondane National Park, Central Norway. Photo: Elisabeth Stur cc-by.

The EBAI project will develop and test new ways of monitoring freshwater invertebrates, and we identify species in environmental samples using short, standardized DNA fragments, so called environmental barcoding. By introducing this cost-effective, rapid and repeatable technique in nature management, monitoring may be expanded also to more remote locations.

One way of doing environmental barcoding is to sample DNA that animals shed to their environment, in water in our case. This environmental DNA (eDNA) can be collected to a filter from the sampled water and be extracted from the filter for downstream analyses. Usually, investigators have carried the sampled water to their laboratory for filtration, but carrying liters of water from remote locations is not practical. Thus, we tested if we could carry some filtration equipment to the field (Fig. 2), and do the filtration in the field (Fig. 3), and then carry just the filters to the lab. Filtering in the field allows more sampling in remote locations and therefore, gives more data for the assessments. The critical part in the in-field filtration approach is how to preserve the filters.

Fig. 2. For filtration in the field, we used an electrical vacuum pump and a manifold with three filter holder bases. We preserved the filters in microcentrifuge tubes or in sterile petri dishes. Photo: Torbjørn Ekrem cc-by.
Fig. 3. We used a car trunk as our filtration laboratory and a bensin-driven aggregate as a power supply for our pump. However, the water filtration is possible using a hand pump if a simpler way is preferred. Photo: Torbjørn Ekrem cc-by.

In our first experiment, we filtered 64 litres of water in the field and tested how different filtration techniques affect the eDNA-based invertebrate community results (Fig. 4).

We found out that mixed cellulose ester filters preserved either dry or in lysis buffer give the most consistent community composition (Fig. 5). Thus, we advocate filtering in the field, using mixed cellulose ester filters and preserving the filters either dry or in lysis buffer. You can find the results summarized in our recently published article:

Fig. 4. Experimental setup. Water samples were collected from two sites, the River Atna and the Lake Jonsvatn. One litre was filtered and eDNA captured onto 0.20-µm polyethersulfone (PES) or 0.45-µm mixed cellulose ester (CN) filters at the river site. At the lake site, eDNA was captured onto 0.45-µm CN filters either directly or after pre-filtration using 12-µm CN filters. Filters were stored in 99% ethanol (EtOH), silica gel (Dry), Qiagen lysis buffer ATL (Buffer) or kept cold (Ice) until DNA was extracted in the laboratory. 500 mL of molecular grade water was filtered and the filters stored with the respective methods as negative controls (B). Fig. 1 in Majaneva et al. 2018 Sci Rep 8: 4682.
Fig. 5. Similarity of community composition in River Atna (a,b) and in Lake Jonsvatn (c,d) samples. The solid line gives the mean similarity and the dashed lines give the 95% confidence intervals. If the 95 % confidence intervals overlap, there is no difference in the methods. The small letters denote significantly different groupings of treatments. Fig. 4 in Majaneva et al. 2018 Sci Rep 8: 4682.

One of the goals of the EBAI project is to distribute knowledge on DNA metabarcoding. We started with a small workshop with two students: Zhenhua Sun from Chalmers University, Sweden and Xiaolong Lin from NTNU University Museum. Our goal was to go through the steps of sample processing using own datasets.

To have a proper feeling on the methodology, we went to sample some water for eDNA analyses in Theissendammen. A suitable outfit for this kind of near-city sampling is a smart casual. (Photos Xiaolong Lin)

We expect to find mallard DNA in our samples. Also beaver is quite potential finding as Markus steps over a beaver dinner to go for the sample. (Photos Xiaolong Lin and Erik Boström)

We visited the molecular lab and Markus showed how to extract DNA and amplify your target gene in PCR as well as explained the paired-end sequencing. (Photos Xiaolong Lin)

Using our own laptops, we analyzed our data. (Photo Xiaolong Lin)

This is one of the questions we want to answer in our EBAI sub-project in Bymarka near Trondheim. Sampling was done before and after rotenone treatment in 2016 and we now follow up with three sampling events in the 2017-season. The first was done early June and the second just finished in late July. We sampled the exact same stations as last year: three different substrates in each six lakes. Three of the lakes had been treated with rotenone, the other three are untreated reference lakes.

With our sampling scheme, we also want to compare eDNA communities with the communities recorded by traditional means (kick-sampling), and see if DNA from species found on one particular substrate leaks to nearby sites in the lake where the species are not recorded.

Reporter Guro Tarjem from The Norwegian Broadcasting Corporation (NRK) just recently visited us in Trondheim and made this radio-report from our study (Norwegian only): Hvem overlevde rotenondrapet?

The European beaver (Castor fiber) is definitely present at Haukvatnet. No surprise if we detect its DNA in our water samples! Photo: Torbjørn Ekrem CC-BY.

The kick-off conference for the EU COST-Action DNAqua-Net was held in the beginning of March 2017. Markus and Torbjørn participated with a poster showing our results from the EBAI eDNA filter preservation experiment (see below).

The conference gathered 172 participants from 37 nations and the impressive line-up of invited speakers communicated the their latest research results on topics relevant to the DNAqua-Net main goal: to develop new genetic tools for bioassessment of aquatic ecosystems in Europe.

Read more about all presentations on the DNAqua-Net website.

We received good feedback on our results and were encouraged to publish these as soon as possible. This we plan to do!

EBAI poster in Essen


Results and prospects of the EBAI-project were discussed as members of the consortium met at the NTNU University Museum on November 10, 2016. Presentations were held by Mehrdad Hajibabaei, Kathrin Langen, Ola Diserud, Markus Majaneva and Torbjørn Ekrem and additional participants were Anders Hobæk, Terje Bongard, Elisabeth Stur, Karstein Hårsaker, Gaute Kjærstad and Erik Boström.

It was a very fruitful meeting where the direction of the project was corrected and the way forward fine tuned according to developments in the field of environmental barcoding of freshwater ecosystems.

The day was rounded off with an excellent meal at EmiliesEld:

Dinner at EmiliesEldReady for desserts at EmiliesEld. Photo Torbjørn Ekrem CC-BY.

The next day, some of the participants toured the vicinities of Trondheim to take a closer look at the sample localities. Here we are at the small lake Holstdammen in Bymarka near Trondheim:

Workshop tour at HolstdammenKathrin Langen, Markus Majaneva, Anders Hobæk and Torbjørn Ekrem touring sites in the EBAI project. Here at reference lake Holstdammen close to Trondheim. Photo: Torbjørn Ekrem CC-BY.

lauglovatnetLauglovatnet, one of the reference lakes with early morning fog. Photo Torbjørn Ekrem CC-BY.

What will happen with the eDNA of lake fauna when much of the animal life is eradicated? We suspect that it will increase as the dead animals dissolve and then decrease to a level below that before rotenone treatment, but we don't know for sure. Previous studies also show that macroinvertebrates in lakes are less susceptible to the rotenone poison than fauna in running water. Thus, we do expect a many species and perhaps species groups to do survive the treatment.

What will happen to these populations when the benthic community is altered and all fish is gone? We will follow up with eDNA sampling through EBAI and compare results with those of traditional benthic kick-sampling at the exact same sites.

This week, a journalist and photographer from NRK - the national public broadcaster - joined us in the field. Their report can be viewed here: "Ny metode for vannovervåkning".

gaute_kjaerstad_water-samplingGaute Kjærstad not quite believing that this is sufficient to tell who lives in Lauglovatnet. Photo Torbjørn Ekrem CC-BY.

aina-aspaas_water-samplingAina Mærk Aspaas sampling water from Theisendammen. Photo Torbjørn Ekrem CC-BY.

markus-majaneva_filtering-waterMarkus Majaneva filtering water samples in the lab at NTNU University Museum, Department of Natural History. Photo Torbjørn Ekrem CC-BY.

This week, lakes in the Trondheim city forest are treated with rotenone to remove common roach (Rutilus rutilus). We know that all fish will be killed by rotenone, but previous studies indicate that macroinvertebrates in lakes are less affected than previously thought (Arnekleiv et al. 2015). In EBAI we will use eDNA and DNA barcoding of bulk macroinvertebrate samples to see how molecular tools can be used to monitor changes in the invertebrate communities after rotenone treatment. Will we detect a change in the invertebrate community? How long will it take until changes are detected as eDNA in water samples? Time will show.

treating-kyvannet-with-rotenoneBoats are used to distribute the rotenone evenly throughout the lake Kyvatnet. Photo: Torbjørn Ekrem cc-by.

Several lakes in the Trondheim city forest will soon be treated with rotenone to remove the common roach (Rutilus rutilus), an invasive alien species in our area. Together with freshwater ecologists at the NTNU University Museum, we are interested to see how the macroinvertebrates react to the treatment and if we can use eDNA to monitor changes and shifts in the species compositions.

Yesteday, we sampled three of the lakes that will undergo treatment as well as three reference lakes. We have three sampling stations at each lake and filter 2 x 1 liter of water from each station. This video shows how samples are taken.

Weather was great and fieldwork successful.

lauglovatnet Gaute Kjærstad sampling water at the lake Lauglovatnet. Photo Torbjørn Ekrem CC-BY.

theisendammenTwo water samples from the lake Theisendammen. Photo Torbjørn Ekrem CC-BY.


The Nedre Dørålseter cabin next to the Rondane National Park in Central Norway. Not a bad place to do field work on a day like this!

In year two of field work for the EBAI-project, we have focused on eDNA and benthic samples at three different stations along the River Atna. By filtering water and analysing DNA from bulk extractions, we want to see how different the diversity of freshwater invertebrates is between stations and if both methods retrieve the same species. As usual, we filtered our water in the field and fixed our 4 min kick samples on ethanol.


Torbjørn sampling water in sterile PET bottles. Will one liter of river water reveal the diversity of macroinvertebrates living here?


Markus filtering water in our mobile lab (click here to see a video). Hikers passing were looked suspiciously and probably wondered what kind of dangerous stuff we were doing.


Elisabeth fixing kick-samples of benthic invertebrates. See here for a video of the actual kicksampling.