Prosjektnummer
901092
High throughput eDNA surveys for benthic monitoring of salmon farms in Norway: A validation study
Ny kunnskap om metoder for bunnovervåking under oppdrettsanlegg og derved viktig bidrag til bærekraftig havbruk
Main findings
• The eDNA metabarcoding approach developed in this project is sufficiently mature to be implemented for all soft-sediment sites situated in the areas, where reference datasets are available.
• Meiofaunal taxa, in particular nematodes, have potential to replace the macro-invertebrates as alternative bioindicators of organic enrichment in metabarcoding data.
• Biotic indices predicted from metabarcoding data using machine learning method show similar ecological status as the indices inferred from macro-invertebrates surveys.
Kortfattet sammendrag av resultater fra prosjektets faglige sluttrapport (FHFs oversettelse) (summery in English further below)
Prosjektet hadde som målsetting å utvikle og validere eDNA-metabarcode-testing for rask, sensitiv og kostnadseffektiv bentisk overvåking av lakseoppdrettslokaliteter i Norge. Det ble generert store mengder metabarcode-data, som ved analyse demonstrerte at eDNA kan benyttes som et verktøy for å vurdere økologisk status for bunndyrsamfunnet. Prosjektet har vist viktigheten av meiofauna, særlig nematoder, som svært gode indikatorer for organisk anriking, og belyst begrensningene med å benytte makrofauna-data i metabarcode-analyser. Prosjektet omfattet også bruk av veiledet maskinlæring for å predikere bentiske indekser, og demonstrerte at dette er effektivt og nøyaktig sammenlignet med konvensjonelle taksonomiske metoder.
Vitenskapelig publisering
Se nederst.
Short summary of results from the project’s final reporting
The project aimed at developing and validating eDNA metabarcoding tests for fast, sensitive, and cost-effective benthic monitoring of salmon farms in Norway. It generated huge amount of metabarcoding data, whose analysis demonstrated the usefulness of eDNA metabarcoding as a tool for assessing the ecological status of benthic community. The project showed the importance of meiofauna, in particular nematodes, as excellent indicators of organic enrichment, and highlighted the limitations of using macrofaunal data in metabarcoding analyses. The project also introduced supervised machine learning to predict benthic indices, demonstrating its effectiveness and accuracy compared to conventional taxonomic methods.
Scientific publications
– Tristan Cordier, Anders Lanzén, Laure Apothéloz‐Perret‐Gentil, Thorsten Stoeck, and Jan Pawlowski, ‘Embracing Environmental Genomics and Machine Learning for Routine Biomonitoring’, Trends in Microbiology, 27/5 (2019), 387–97. For an abstract and ordering details, see Flinkinghub.elsevier at https://doi.org/10.1016/j.tim.2018.10.012.
– Jan Pawlowski, Mary Kelly-Quinn, Florian Altermatt, and Maria Kahlert, ‘The future of biotic indices in the ecogenomic era: Integrating (e)DNA metabarcoding in biological assessment of aquatic ecosystems’, Science of the Total Environment, 638 (2018), 1295–1310, https://www.researchgate.net/deref/http%3A%2F%2Fdx.doi.org%2F10.1016%2Fj.scitotenv.2018.05.002 (open access).
– Tristan Cordier and Jan Pawlowski, ‘BBI: an R package for the computation of Benthic Biotic Indices from composition data’, Metabarcoding and Metagenomics, 2/e25649 (2018), https://doi.org/10.3897/mbmg.2.25649 (open access).
– Tristan Cordier, Dominik Forster, Yoann Dufresne, Catarina I. M. Martins, Thorsten Stoeck, and Jan Pawlowski, ‘Supervised machine learning outperforms taxonomy‐based environmental DNA metabarcoding applied to biomonitoring’, Molecular Ecology Resources, 18/6 (2018), 1381– 91. For an abstract and ordering details, see Wiley Online Library at https://doi.org/10.1111/1755-0998.12926.
– Thorsten Stoeck, Larissa Frühe, Dominik Forster, Tristan Cordier, Catarina I. M. Martins, and Jan Pawlowski, ‘Environmental DNA metabarcoding of benthic bacterial communities indicates the benthic footprint of salmon aquaculture’, Marine Pollution Bulletin, 127 (2018), 139-49. For abstract ordering details, see ScienceDirect at https://doi.org/10.1016/j.marpolbul.2017.11.065.
– Tristan Cordier, Philippe Esling, Franck Lejzerowicz, Joana Visco, Amine Ouadahi, Catarina Martina, Tomas Cedhagen, and Jan Pawlowski, ‘Predicting the Ecological Quality Status of Marine Environments from eDNA Metabarcoding Data Using Supervised Machine Learning’, Environmental Science & Technology, 51/16 (2017), 9118–26. For ordering details, see ACS Publications at https://doi.org/10.1021/acs.est.7b01518.
The project aimed at developing and validating eDNA metabarcoding tests for fast, sensitive, and cost-effective benthic monitoring of salmon farms in Norway. It generated huge amount of metabarcoding data, whose analysis demonstrated the usefulness of eDNA metabarcoding as a tool for assessing the ecological status of benthic community. The project showed the importance of meiofauna, in particular nematodes, as excellent indicators of organic enrichment, and highlighted the limitations of using macrofaunal data in metabarcoding analyses. The project also introduced supervised machine learning to predict benthic indices, demonstrating its effectiveness and accuracy compared to conventional taxonomic methods.
Scientific publications
– Tristan Cordier, Anders Lanzén, Laure Apothéloz‐Perret‐Gentil, Thorsten Stoeck, and Jan Pawlowski, ‘Embracing Environmental Genomics and Machine Learning for Routine Biomonitoring’, Trends in Microbiology, 27/5 (2019), 387–97. For an abstract and ordering details, see Flinkinghub.elsevier at https://doi.org/10.1016/j.tim.2018.10.012.
– Jan Pawlowski, Mary Kelly-Quinn, Florian Altermatt, and Maria Kahlert, ‘The future of biotic indices in the ecogenomic era: Integrating (e)DNA metabarcoding in biological assessment of aquatic ecosystems’, Science of the Total Environment, 638 (2018), 1295–1310, https://www.researchgate.net/deref/http%3A%2F%2Fdx.doi.org%2F10.1016%2Fj.scitotenv.2018.05.002 (open access).
– Tristan Cordier and Jan Pawlowski, ‘BBI: an R package for the computation of Benthic Biotic Indices from composition data’, Metabarcoding and Metagenomics, 2/e25649 (2018), https://doi.org/10.3897/mbmg.2.25649 (open access).
– Tristan Cordier, Dominik Forster, Yoann Dufresne, Catarina I. M. Martins, Thorsten Stoeck, and Jan Pawlowski, ‘Supervised machine learning outperforms taxonomy‐based environmental DNA metabarcoding applied to biomonitoring’, Molecular Ecology Resources, 18/6 (2018), 1381– 91. For an abstract and ordering details, see Wiley Online Library at https://doi.org/10.1111/1755-0998.12926.
– Thorsten Stoeck, Larissa Frühe, Dominik Forster, Tristan Cordier, Catarina I. M. Martins, and Jan Pawlowski, ‘Environmental DNA metabarcoding of benthic bacterial communities indicates the benthic footprint of salmon aquaculture’, Marine Pollution Bulletin, 127 (2018), 139-49. For abstract ordering details, see ScienceDirect at https://doi.org/10.1016/j.marpolbul.2017.11.065.
– Tristan Cordier, Philippe Esling, Franck Lejzerowicz, Joana Visco, Amine Ouadahi, Catarina Martina, Tomas Cedhagen, and Jan Pawlowski, ‘Predicting the Ecological Quality Status of Marine Environments from eDNA Metabarcoding Data Using Supervised Machine Learning’, Environmental Science & Technology, 51/16 (2017), 9118–26. For ordering details, see ACS Publications at https://doi.org/10.1021/acs.est.7b01518.
– Jan Pawlowski, Philippe Esling, Franck Lejzerowicz, Tristan Cordier, Joana A. Visco, Catarina I. M. Martins, Arne Kvalvik, Knut Staven, Tomas Cedhagen, ‘Benthic monitoring of salmon farms in Norway using foraminiferal metabarcoding’,
Aquaculture Environmental Interaction, 8 (2016), 371–86,
https://doi.org/10.3354/aei00182 (open access).
Prosjektet har vist at eDNA-metabarcoding er et nytt verktøy for overvåking av tilstanden i bunnsedimenter under oppdrettsanlegg. Metoden er raskere og rimeligere enn dagens metoder, og vil med stor sannsynlighet kunne utvikles slik at den kan erstatte konvensjonelle, tidkrevende metoder med bunndyrtelling. Sannsynligvis gir metoden mer informasjon om bunntilstanden enn dagens tradisjonelle metoder, men det er noen utfordringer knyttet til at de ikke alltid gir samme “svar” på bunntilstanden. Når de to metodene er benyttet flere steder parallelt vil dette antagelig gi kunnskapsgrunnlag som gjør at overvåking kan skje med den nye, raske og effektive metoden eDNA metabarcoding.
-
Fact sheet: Benthic monitoring of salmon farms in Norway using environmental DNA metabarcoding
University of Geneva, Switzerland; ID-Gene Ecodiagnos Ltd, Plan-les-Ouates, Switzerland; and University of Aarhus, Denmark. Poster presented at the Aquaculture Europe 2016 conference. September 2016. By Jan Pawlowski (Department of Genetics and Evolution, University of Geneva and ID-Gene Ecodiagnostics Ltd), Trisdan Cordier (Department of Genetics and Evolution, University of Geneva), Philippe Esling (Department of Genetics and Evolution, University of Geneva), Frank Lejzerowicz (Department of Genetics and Evolution, University of Geneva), Joana Visco (ID-Gene Ecodiagnostics Ltd), and Tomas Cedhagen (Department of Biological Sciences).
-
Final report: High throughput eDNA surveys for benthic monitoring of salmon farms in Norway: a validation study
Université de Geneve, ID-Gene ecodiagnostics. Report. 30 October 2019. By Jan Pawlowski (UNIGE/ID-Gene ecodiagnostics), Tristan Cordier (UNIGE/ID-Gene ecodiagnostics), Kristina Cermakova (UNIGE/ID-Gene ecodiagnostics), and Tomas Cedhagen (University of Aarhus).
-
Note: Guidelines for using environmental DNA (eDNA) to monitor the impact of salmon farmson benthic communities
Université de Geneve / ID-GENE ecodiagnostics. 29. March 2019. By Jan Pawlowski, Trisdan Cordier, and Kristina Cermakova.
-
Occasional paper: High-throughput eDNA surveys for benthic monitoring ofsalmon farms in Norway: a validation study
Draft report from scientific panel. 10 September 2018. Av Thorsten Stoeck (UoKaiserlautern – Germany), Anders Lanzen (AZTI – Spain), Thierry Baussant (NORCE Research – Norway), Trisdan Cordier (UoGeneve – Switzerland), and Jan Pawlowski (UoGeneve – Switzerland).
Gjennomføring av bunnundersøkelser i henhold til MOM-C (en bløtbunnundersøkelse som inngår i MOM–systemet (Matfiskanlegg – Overvaking – Modellering), med telling og klassifisering av bunndyr, er et tidkrevende og kostbart arbeid. Utviklingen av nye DNA-baserte metoder (Next Generation Sequencing – NGS) har åpnet muligheten for å bruke slik metodikk for å karakterisere bunndyrsamfunn, og gjennom det dokumentere påvirkningen fra oppdrettsanlegg. Det er gjennomført flere foreløpige studier der de nye metodene er tatt i bruk, med lovende resultater. Fremdeles gjenstår en del metodeutvikling og verifisering for norske forhold før det vil være mulig å erstatte de tidligere metodene med nye, raskere, nøyaktigere og antagelig rimeligere metoder.
Background
Recent development of NGS technologies offers the possibility to use environmental DNA or RNA to explore benthic diversity. The NGS-based approach, also called metabarcoding, has been proposed as a cost effective way to overcome the limitations of morphological identification in routine biomonitoring, allowing the processing of large numbers of samples for increasingly efficient and reliable surveys
The implementation of NGS approach to the routine benthic monitoring is currently limited by a relatively low efficiency in detecting macrofaunal species in small volumes of sediment used for eDNA studies. To increase the accuracy of this approach it would be necessary to complement the study of macrofaunal bio-indicators with the analysis of meiofauna species, including small-size metazoans (nematodes, copepods) and foraminifera.
A pilot study by the project team using NGS metabarcoding for environmental monitoring confirmed the potential of some foraminiferal taxa as indicators of enrichment associated with the salmon farming activities in Scotland. Similar results have been obtained for the foraminifera-oriented monitoring studies of salmon farms in New Zealand and Norway. At the same time, the project team has analysed the biotic indices inferred from metazoan eDNA data. The team`s studies in Scotland and Norway shows that the NGS metazoan data provide very similar evaluation of environmental impact as the morpho-taxonomic analyses. However, the conclusions of these studies also highlight the necessity of including meiofauna as potential bio-indicators in the future.
The present project aims at validating these preliminary results by analysing more sites representing different ecological conditions. These validation studies will focus on two particularly important aspects: the detection of macrofaunal species and the identification of meiofaunal bio-indicators. The project will also address the need for adapting the biotic indices used in benthic monitoring to the specificities of NGS data.
The implementation of NGS approach to the routine benthic monitoring is currently limited by a relatively low efficiency in detecting macrofaunal species in small volumes of sediment used for eDNA studies. To increase the accuracy of this approach it would be necessary to complement the study of macrofaunal bio-indicators with the analysis of meiofauna species, including small-size metazoans (nematodes, copepods) and foraminifera.
A pilot study by the project team using NGS metabarcoding for environmental monitoring confirmed the potential of some foraminiferal taxa as indicators of enrichment associated with the salmon farming activities in Scotland. Similar results have been obtained for the foraminifera-oriented monitoring studies of salmon farms in New Zealand and Norway. At the same time, the project team has analysed the biotic indices inferred from metazoan eDNA data. The team`s studies in Scotland and Norway shows that the NGS metazoan data provide very similar evaluation of environmental impact as the morpho-taxonomic analyses. However, the conclusions of these studies also highlight the necessity of including meiofauna as potential bio-indicators in the future.
The present project aims at validating these preliminary results by analysing more sites representing different ecological conditions. These validation studies will focus on two particularly important aspects: the detection of macrofaunal species and the identification of meiofaunal bio-indicators. The project will also address the need for adapting the biotic indices used in benthic monitoring to the specificities of NGS data.
Å utvikle NGS eDNA-undersøkelser til et raskt, sensitivt og kost-effektivt verktøy for bunnundersøkelser under lakseoppdrettsanlegg i Norge.
Objectives
To develop the NGS eDNA surveys for a fast, sensitive, and cost-effective benthic monitoring of salmon farms in Norway.
Sub-objectives
To develop the NGS eDNA surveys for a fast, sensitive, and cost-effective benthic monitoring of salmon farms in Norway.
Sub-objectives
The project will comprise three tasks, which specific aims are:
1. to improve the detection of benthic macroinvertebrates in eDNA samples;
2. to identify and validate new meiofaunal bioindicators;
3. to develop and validate the NGS-based biotic indices combining the macro- and meiofaunal data.
1. to improve the detection of benthic macroinvertebrates in eDNA samples;
2. to identify and validate new meiofaunal bioindicators;
3. to develop and validate the NGS-based biotic indices combining the macro- and meiofaunal data.
For næringen vil nytteverdien være å:
• redusere tiden for prøveanalyser
• redusere kostnader for prøvetaking
• redusere kostnader for analysearbeidet
• redusere kostnader for prøvetaking
• redusere kostnader for analysearbeidet
Sammenlignet med dagens metoder kan det forventes raskere resultater (2 uker mot dagens mer enn 3 måneder), og en halvering av kostnadene.
Som verktøy for miljøovervåking vil metoden:
• være mer følsom enn dagens metoder
• være mer nøyaktig siden flere arter inkluderes
• bidra til økt standardisering
Expected project impact
The development of NGS eDNA surveys will modernize the benthic monitoring ensuring the sustainable growth of the industries.
From the economic point of view, the main advantages of NGS approach consist in:
• reducing time of sample analyses allowing much faster evaluation of the salmon farms impact;
• reducing costs of sampling (no sieving, no additional costs for fixed sample storage and/or toxic products transportation and disposal);
• reducing costs of data acquirement and analyses by automatized workflow.
From the point of view of environmental management, the NGS surveys will allow:
• increasing sensitivity of environmental monitoring by expanding the range of bio-indicators;
• increasing accuracy by more reliable species identification (including tissue fragments and immature stages);
• increasing uniformity of data analyses by using standardized DNA barcodes for species identification and application of standardized bioinformatics tools.
Compared to the currently used morpho-taxonomic approach, the NGS surveys should provide the results much more rapidly (2 weeks instead of > 3 months) and at lower price (150–200 € instead of > 350 € per sample).
The development of NGS eDNA surveys will modernize the benthic monitoring ensuring the sustainable growth of the industries.
From the economic point of view, the main advantages of NGS approach consist in:
• reducing time of sample analyses allowing much faster evaluation of the salmon farms impact;
• reducing costs of sampling (no sieving, no additional costs for fixed sample storage and/or toxic products transportation and disposal);
• reducing costs of data acquirement and analyses by automatized workflow.
From the point of view of environmental management, the NGS surveys will allow:
• increasing sensitivity of environmental monitoring by expanding the range of bio-indicators;
• increasing accuracy by more reliable species identification (including tissue fragments and immature stages);
• increasing uniformity of data analyses by using standardized DNA barcodes for species identification and application of standardized bioinformatics tools.
Compared to the currently used morpho-taxonomic approach, the NGS surveys should provide the results much more rapidly (2 weeks instead of > 3 months) and at lower price (150–200 € instead of > 350 € per sample).
Project design and implementation
The project comprises the following three work packages (WP), which address the main issues revealed by the preliminary studies:
WP 1. NGS analysis of benthic macrofauna
Objectives: This WP aims at improving the detection of macro-invertebrate species in eDNA samples. This aim will be achieved by:
• WP1.1 – enriching DNA barcodes database to better assess genetic diversity of coastal macro-invertebrates;
• WP1.2 – testing alternative sampling protocols to increase the representativeness of macrofaunal species in eDNA samples.
WP 2. Identification and validation of meiofauna bioindicators
Objectives: The aim of this WP is to search for meiofaunal species that could complement the set of macrofaunal indicators. To fulfil this aim, the team will analyse the distribution of metazoan meiofauna and foraminifera present in the NGS data and their variations along environmental gradients. Special emphasis will be given to testing the micropatchiness of meiofaunal species in order to determine accurate sampling conditions. The selected candidate species will be compared to the project team's local NGS database and those showing a consistent affinity to certain environmental conditions will be assigned with corresponding ecological values.
WP 3. Revision of benthic community indices
The project comprises the following three work packages (WP), which address the main issues revealed by the preliminary studies:
WP 1. NGS analysis of benthic macrofauna
Objectives: This WP aims at improving the detection of macro-invertebrate species in eDNA samples. This aim will be achieved by:
• WP1.1 – enriching DNA barcodes database to better assess genetic diversity of coastal macro-invertebrates;
• WP1.2 – testing alternative sampling protocols to increase the representativeness of macrofaunal species in eDNA samples.
WP 2. Identification and validation of meiofauna bioindicators
Objectives: The aim of this WP is to search for meiofaunal species that could complement the set of macrofaunal indicators. To fulfil this aim, the team will analyse the distribution of metazoan meiofauna and foraminifera present in the NGS data and their variations along environmental gradients. Special emphasis will be given to testing the micropatchiness of meiofaunal species in order to determine accurate sampling conditions. The selected candidate species will be compared to the project team's local NGS database and those showing a consistent affinity to certain environmental conditions will be assigned with corresponding ecological values.
WP 3. Revision of benthic community indices
Objective: This WP will seek to modify currently used biotic indices in benthic monitoring. The project team will use the NGS eDNA data obtained in WP 1 and WP 2 in order to devise a novel index that will combine macro- and meiofaunal data and will fit to the specificities of the NGS approach.
Forlengelse / extension 2018
An evaluation of result by the project and reference group by the end of 2017 concluded that an extension period will allow wrapping up the project by reaching the following goals:
• to complete the sampling and genetic analysis of some additional sites selected based on their geographic position and privileging the under-sampled habitats;
• to barcode and morphologically characterize the potential bioindicator meiofauna species that commonly appear in eDNA data;
• to test the robustness of machine learning and other statistical methods used for analysis of eDNA data;
• to prepare the guidelines for standardized protocols.
In addition, this extension period will allow creating a group of international experts that will critically assess the accuracy and robustness of machine learning and other statistic methods used for the analyses of eDNA data.
• to complete the sampling and genetic analysis of some additional sites selected based on their geographic position and privileging the under-sampled habitats;
• to barcode and morphologically characterize the potential bioindicator meiofauna species that commonly appear in eDNA data;
• to test the robustness of machine learning and other statistical methods used for analysis of eDNA data;
• to prepare the guidelines for standardized protocols.
In addition, this extension period will allow creating a group of international experts that will critically assess the accuracy and robustness of machine learning and other statistic methods used for the analyses of eDNA data.
Dissemination of project results
• Papers published in peer-reviewed scientific journals on:
o meiofaunal bioindicators in benthic monitoring
o NGS analysis of benthic macrofauna in ethanol fixed samples
o revision of benthic indices for NGS data;
• Presentations at international conferences (oral talks, posters);
• Project report to FHF assessing the validity of the NGS eDNA surveys.
• Papers published in peer-reviewed scientific journals on:
o meiofaunal bioindicators in benthic monitoring
o NGS analysis of benthic macrofauna in ethanol fixed samples
o revision of benthic indices for NGS data;
• Presentations at international conferences (oral talks, posters);
• Project report to FHF assessing the validity of the NGS eDNA surveys.
-
Final report: High throughput eDNA surveys for benthic monitoring of salmon farms in Norway: a validation study
Université de Geneve, ID-Gene ecodiagnostics. Report. 30 October 2019. By Jan Pawlowski (UNIGE/ID-Gene ecodiagnostics), Tristan Cordier (UNIGE/ID-Gene ecodiagnostics), Kristina Cermakova (UNIGE/ID-Gene ecodiagnostics), and Tomas Cedhagen (University of Aarhus).