Forum Paper |
Corresponding author: Rosetta C. Blackman ( rosiecblackman@gmail.com ) Academic editor: Dirk Steinke
© 2019 Rosetta C. Blackman, Elvira Mächler, Florian Altermatt, Amanda Arnold, Pedro Beja, Pieter Boets, Bastian Egeter, Vasco Elbrecht, Ana Filipa Filipe, J. Iwan Jones, Jan Macher, Markus Majaneva, Filipa M. S. Martins, Cesc Múrria, Kristian Meissner, Jan Pawlowski, Paul L. Schmidt Yáñez, Vera M.A. Zizka, Florian Leese, Benjamin Price, Kristy Deiner.
This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Citation:
Blackman RC, Mächler E, Altermatt F, Arnold A, Beja P, Boets P, Egeter B, Elbrecht V, Filipe AF, Jones JI, Macher J, Majaneva M, Martins FMS, Múrria C, Meissner K, Pawlowski J, Schmidt Yáñez PL, Zizka VMA, Leese F, Price B, Deiner K (2019) Advancing the use of molecular methods for routine freshwater macroinvertebrate biomonitoring – the need for calibration experiments. Metabarcoding and Metagenomics 3: e34735. https://doi.org/10.3897/mbmg.3.34735
|
Over the last decade, steady advancements have been made in the use of DNA-based methods for detection of species in a wide range of ecosystems. This progress has culminated in molecular monitoring methods being employed for the detection of several species for enforceable management purposes of endangered, invasive, and illegally harvested species worldwide. However, the routine application of DNA-based methods to monitor whole communities (typically a metabarcoding approach) in order to assess the status of ecosystems continues to be limited. In aquatic ecosystems, the limited use is particularly true for macroinvertebrate communities. As part of the DNAqua-Net consortium, a structured discussion was initiated with the aim to identify potential molecular methods for freshwater macroinvertebrate community assessment and identify important knowledge gaps for their routine application. We focus on three complementary DNA sources that can be metabarcoded: 1) DNA from homogenised samples (bulk DNA), 2) DNA extracted from sample preservative (fixative DNA), and 3) environmental DNA (eDNA) from water or sediment. We provide a brief overview of metabarcoding macroinvertebrate communities from each DNA source and identify challenges for their application to routine monitoring. To advance the utilisation of DNA-based monitoring for macroinvertebrates, we propose an experimental design template for a series of methodological calibration tests. The template compares sources of DNA with the goal of identifying the effects of molecular processing steps on precision and accuracy. Furthermore, the same samples will be morphologically analysed, which will enable the benchmarking of molecular to traditional processing approaches. In doing so we hope to highlight pathways for the development of DNA-based methods for the monitoring of freshwater macroinvertebrates.
Bulk DNA, community, DNAqua-Net, environmental DNA, experimental methods, fixative DNA, monitoring, Water Framework Directive
Worldwide, DNA-based methods are advancing and can aid in the determination of ecological state of ecosystems. In Europe, the COST Action DNAqua-Net, consisting of more than 500 members and is working to utilise and improve molecular methods for monitoring Biological Quality Elements (BQEs, e.g. fish, macroinvertebrates, and phytoplankton-benthos) used to determine aquatic ecosystem status under the requirements of the Water Framework Directive (WFD, 2000/60/EC) and beyond (
On the 17th December 2018, members of the DNAqua-Net consortium met at Centro de Investigação em Biodiversidade e Recursos Genéticos (CIBIO), Portugal, and discussed future developments of DNA-based methods for monitoring macroinvertebrates. Based on discussions and collating information from members within the COST Action, knowledge gaps were identified, and research priorities outlined focusing on two key areas: (1) current research needs and the short-term focus of method development and (2) issues not being addressed by current official monitoring methods and the potential for molecular methods to address these. In this article, we summarise the outcomes of these discussions, including areas of consensus, knowledge gaps, proposed experimental designs to fill those knowledge gaps, and future opportunities for molecular approaches to aid aquatic bioassessment using macroinvertebrates.
There are three main sources of DNA from macroinvertebrates being assessed for use in biomonitoring programs (Fig.
The three methods for DNA retrieval for macroinvertebrate biomonitoring: (1) DNA bulk samples from homogenised specimen samples, (2) DNA extracted from the fixative used to preserve or store a sample, and (3) DNA extracted from a water sample (eDNA).
To retain comparability with current practices, but utilise the benefits of DNA-based methods, a consensus emerged that for the time being, DNA from a bulk sample and fixative DNA are the most comparable options to current methods used for applied biomonitoring of freshwater macroinvertebrates. These methods rely on the same field sampling methods used under current official monitoring practices and have the same spatial interpretation. The advantage of using DNA metabarcoding from sources (1) and (2) compared with current morphological identification of samples is the potential for increased time efficiency and identification resolution. Although not yet demonstrated with current bulk DNA processing methods, in the case of fixative DNA, removal of the time intensive step needed to sort the macroinvertebrate sample from organic material collected in the sampling process, while also avoiding destruction of the species is particularly important. However, both methods do not overcome possible limitations encompassed by the physical sampling of organisms, and this is where developing eDNA methods may be utilised (see Future Application for further discussion).
We identified several gaps in our current understanding warranting further research. Primarily, research has focused on small scale comparisons within single systems or countries. This form of testing does not allow for variation in geographic range or among water quality classes and should be included within future experimental design. Secondly, current macroinvertebrate community indices require abundance or frequency classes for community assessment. Metabarcoding studies have so far been unable to relate read number precisely to abundance, relative abundance or biomass of macroinvertebrates. A recent study by
A key advancement for any new method to be adopted in EQS assessment requires a reduction in sample processing cost (both time and monetary); an issue repeatedly raised by regulators during DNAqua-Net stakeholder meetings. Recent studies using bulk DNA samples often include specimens being picked from the sample matrix, a form of size sorting of specimens, or removing legs from individuals prior to processing (
With the goal to develop robust protocols for macroinvertebrate assessment from DNA extracted from samples collected using current official sampling methods, we outline a set of experiments (Fig.
Outline of the proposed intercalibration experiments: A. Pre-extraction: Each lab will process one standard WFD macroinvertebrate sample to send to the other four labs. Each lab will work with five fixative samples and five blended bulk samples, four of which are unknown. Each lab will run their custom lab and bioinformatic pipelines and an agreed standard pipeline for the extraction and downstream analysis. B. Post-extraction: Each lab will create one macroinvertebrate bulk DNA mock sample to send to the other four labs, thus each lab will work with five mock community samples, four of which are unknown to analyse. Each lab will run their custom lab and bioinformatic pipelines and an agreed standard pipeline.
The performance and comparability of bulk and fixative DNA metabarcoding should be assessed based on samples collected with an established sampling method, such as the multi-habitat kick-net sample (
The DNA source is not the only factor determining variation in results, a key component of variance arises from differing laboratory protocols used to amplify and sequence the DNA. Laboratories use a wide variety of approaches for DNA amplification, PCR primers, library preparation, and bioinformatic pipelines. We therefore propose to establish a set of “blind mock community” samples, each to be tested by contributing labs (Fig.
It is important for the final assessment of pre- and post-extraction variation among the protocols that all data is analysed using the same bioinformatic pipeline to avoid differences due to pre-filtering of sequence reads, clustering algorithm, or taxa assignment. A standard protocol will be decided upon prior to the start of the experiments. However, the data generated from our calibration of methods will lend itself well to further tests of how results in DNA metabarcoding protocols and DNA sources interact with bioinformatic processing decisions. Data generated from these experimental designs will be archived with full metadata to allow for such bioinformatic comparisons.
Over the past century, the response of macroinvertebrates to pollution has been well studied (
Over the next year, the DNAqua-Net community has a distinct advantage to make use of its resources and plans to organise large-scale natural experimental studies, documenting ecological community responses in multi-stressor environments. Importantly, the network will focus on the goal of developing DNA-based methods for monitoring macroinvertebrates and compare them with morphometric identification methods based on macroinvertebrate sampling with kick-nets, (i.e. method alignment). However, we will also look to develop a monitoring approach which differs from the traditional approaches to sample collection (i.e. method independency). The former may allow us to link existing data but may be guided and optimised by past limitations inherent in the traditional approach. While the latter may make comparisons with past samples impossible, it will start without the historic constraints of macroinvertebrate indices. For example, using either DNA from bulk samples or fixative DNA still harbours the limitations of traditional sample methods and is driven by the restrictions caused by the traditional and invasive collection of organisms (i.e. missing locally low abundant taxa including rare, elusive, or invasive species), while this could be bypassed by alternative approaches, such as eDNA.
The use and application of eDNA may be able to resolve these latter issues in part, but is, as a method, not yet standardised for sampling or processing and may give a more complementary measure when compared with the classic macroinvertebrate sampling. The use of eDNA for community detection has rapidly developed in recent years and has been successfully applied to a number of groups, notably fish (
This article has aimed to highlight the status of macroinvertebrate community analysis via DNA-based methods. It is our hope that the experiments designed as part of our discussions and workshop provide immediate areas of research to be undertaken. By carrying out comparison of workflows both within (in-house vs standardised) and across laboratories we will establish key points of the methods which influence the results and we will be able to form a basis for best practice. Bulk DNA samples have been the focus of efforts thus far, and shortly will be looked at on a large geographic scale (e.g. by SCANDNAnet, a project funded by the Nordic Council of Ministers; Joint Danube Survey 4; GeDNA, a project funded by the German Federal Environmental Agency [FKZ 3719242040]; and FRESHING, funded by FCT-Portugal together with EnvMetaGen funded by H2020). However, implementation using this DNA source still requires further assessment at the individual lab scale and streamlining of sample processing methods. Fixative DNA, a promising DNA source, remains largely unexplored and should be further compared with bulk DNA and current monitoring methods. The chances of future uptake of the described molecular methods by regulators into official, mandatory routine monitoring programs such as the WFD and MSFD will be greatly increased by conducting large experimental validation studies and by agreeing on standardised procedural protocols amongst scientists. We encourage the exploration and research of whole catchment-based approaches (via eDNA, including its degradation and transport) and working towards gaining an understanding of macroinvertebrate community responses to new and varied pollutants. For example, eDNA is currently being sampled and analysed in parallel with large ongoing aquatic monitoring programs in Switzerland (
This article is a direct outcome of the European Cooperation in Science and Technology (COST) Action CA15219 DNAqua-Net. RB, FA, and EM are supported by Swiss National Science Foundation Grants No PP00P3_179089 and 31003A_173074 and the University of Zurich Research Priority Programme “URPP Global Change and Biodiversity” (to FA). FA and EM are also supported by the Velux Foundation. PB was supported by EDP-Biodiversity Chair (EDP/FCT). BE is supported by the European Union’s Horizon 2020 research and innovation programme under grant agreement No 668981. VE is supported by funding through the Canada First Research Excellence Fund. AFF is supported by the FRESHING project funded by FCT (Portugal) and COMPETE (PTDC/AAGMAA/2261/2014). MM is supported by the Research Council of Norway (project no. 243791/E50) and the Norwegian Environment Agency (project no. 15040013). FMSM is supported by FCT PhD grant SFRH/BD/104703/2014. KM is supported by BONUS FUMARI; BONUS (art. 185), which is jointly funded by the European Union, the Academy of Finland, and the Swedish Research Council Formas and by the Nordic Council of Ministers to SCANDNAnet, grant no.18103. JP is supported by the Swiss National Science Foundation grant (31003A_179125). FL and VZ are supported by the German Barcode of Life project (GBOL) funded through the German Federal Ministry of Education and Research. We thank Jörg Strackbein (University of Duisburg-Essen) for generating the figures and Sandra Aresta funded by the European Union’s Horizon 2020 research and innovation programme under grant agreement No 668981 (ERA Chair in Environmental metagenomics, EnvMetaGen) and colleagues at CIBIO for facilitating our meeting. We also thank all those that participated in discussions both online and at the meeting. We thank Alex Bush, Erik Pilgrim, and an anonymous reviewer for comments on the manuscript.