Applied Environmental Genomics

Introduction to applied environmental genomics

THE CHALLENGE OF ENVIRONMENTAL MANAGEMENT

The economy is a wholly owned subsidiary of the environment, not the reverse.

— Herman E. Daly

Human life depends entirely on the goods and services provided by Earth’s biological resources. The biosphere is our habitat, providing us the air we breathe, the plants and animals we eat, the fibres that clothe us, the clean water we drink, the pharmaceuticals that maintain our health, and the spiritual sustenance we experience in nature. Ironically, in the context of global climate change and its catastrophic impacts on biodiversity, even our carbon-based fuels have a biological origin. Economists estimate the value of global environmental services to be 125 trillion dollars annually (Almond et al. 2020). This sum is so vast it obscures meaning, but the reality is simple. Without a functioning environment, our economies, our societies, our cultures, and even our bodies cannot function.

It follows that purely from a self-interested perspective, care for the environment must be a core human responsibility since we control the biosphere more than any other species (Bar-On et al. 2018). Of course, many would argue that nature deserves respect and conservation independent of human interests. And yet, daily reminders point to virtually all indicators of environmental function going in the wrong direction. What is to be done?

Policies, efforts and actions - at every level - will only succeed ... when based on the best knowledge and evidence.

— Sir Robert Watson, Chair IPBES

Herein lies the central problem in managing ecosystems on their vast scales: a lack of information to support decision making. This situation is so familiar that it is regarded as unremarkable and is wholly normalised in scientific and political circles. Professor Graham Samuel in his recent critique of the fitness of Australia’s Environmental Protection Biodiversity Conservation Act writes: ‘Decision-makers, proponents and the community do not have access to the best available data, information and science’ (Samuel 2020). In a nutshell, we rarely have information at the scale, speed, and accuracy to make good decisions.

THE RELEVANCE OF GENOMICS

What is the relevance of genomics to this problem? We argue that genomics is relevant because it will be a part of the solution. Genomes encode biological identity, function and condition. Genomes and their products determine what organisms do, from the unique metabolic capabilities of microbes to the behaviour and interactions of large organisms with their environment. They regulate all manner of physiological responses to environmental stimuli. And finally, all levels of biological organisation can be distinguished by their unique DNA. Cells can be identified by transcriptional or epigenomic profiles, individuals can be identified by their genotype, and communities can be identified by the genetic signatures of the species within them. All these aspects – function, condition, and identity – have applications in environmental management. Our challenge is to access this information and make sense of it.

This challenge is within our capabilities. Capitalising on the rich information that genomes provide through advances in molecular biology, information sciences, and engineering, has transformed both medicine and agriculture in recent decades. Genomics has provided access to knowledge to treat human diseases, increase crop yields and more. A similar revolution in environmental science is possible. The massive budgets and engines of innovation in medicine and agriculture have fuelled the development of genomic technologies. The products of this are ripe for co-opting to environmental applications. The chapters of this book show that environmental genomics sits at a unique convergence where the need for better environmental management informed by science is very high, and mature genomic technologies are poised to provide this information.

Of course, the transfer of genomics technologies between medicine, agriculture, and environmental science builds on strong foundations, established over many decades. The methods used for genetic mark-recapture (Chapter 10) are equivalent to those used by forensic scientists for trace DNA collection. The direct genetic modification method provided by CRISPR-Cas9 technology was developed for medical applications but has environmental applications such as gene drives for population control of invasive animal species (Chapter 26) and the real-time monitoring or manipulation of population sex ratios in food production industries (Chapter 8). During the past few years, we have observed the applications of DNA and genomics to environmental science grow rapidly and diversify enormously (Fig. 1). In this book, we set out to capture a greater diversity of applications than typically are included in advanced texts by including applications scattered through the scientific journal literature. We present this exciting field in the words of the scientists leading this technological transformation, and the end users who are adopting new approaches to understanding and managing nature.

Fig. 1. Technological development in environmental genomics. Significant innovations in fields that contribute to environmental genomics are shown. Chapters in this book that relate to each field are indicated. (A) Nucleic acid extraction methods were initially for large samples from single species, but polymerase chain reaction (PCR) based methods enabled ‘trace’ DNA sampling such as the first ‘ancient DNA’ study identifying a museum skin as an extinct Qagga species (Higuchi et al. 1984); and the first dietary analysis of food items from bear faeces in 1988 (Höss et al. 1992). eDNA sampling is the main field method for many chapters in this book, and new developments such as passive eDNA collection are transforming its capacity (Bessey et al. 2021). (B) Genotyping individuals was once generally done with morphological traits or allozymes until PCR enabled amplification of selected DNA regions. Microsatellite loci and microsatellite DNA sequencing were a mainstay of this area until High-throughput sequencing (HTS) was available. Methods for Single nucleotide polymorphism (SNP) genotyping such as ddRAD became cheap and powerful once HTS was available (Taberlet et al. 1999). (C) Species identification is now routinely DNA sequence based, with DNA barcode identification of samples from single species being commonplace. (D) Analysis of species diversity with DNA changed dramatically once HTS was available, as this allows mixed DNA pools to be sequenced with ease. When combined with PCR amplification of a barcode, this led to DNA metabarcoding (Taberlet et al. 2012). DNA metabarcoding replaced clone library analysis and methods based on DNA sequence conformation like DGGE and similar approaches. (E) Genetic modification of bacterial plasmids allowed production of bulk DNA before PCR, and bulk proteins if expression plasmids were constructed. Specific genomic modification methods were developed for a range of eukaryotes such as Saccharomyces, Drosophila, and multiple crop species. Gene knockout methods such as RNA interference technology worked on a wider range of species, but the more recent CRISPR-Cas9 technology provides a general approach for targeted gene modification in most organisms (Bak et al. 2018). (F) Molecular biology underpins all the research applications described in this book. The PCR transformed manipulation and detection of nucleic acids, especially once thermostable polymerases like Taq could be used. (G) The discovery of ‘Archaea’ by phylogenetic analysis of ribosomal RNA sequences overhauled our view of earth’s biodiversity (Woese and Fox 1977). Phylogenomics became possible once large numbers of genome assemblies became available, and bioinformatic methods for selecting phylogenetically informative regions for comparison were developed. Affordable HTS now allows phylogenetic assessment of community uniqueness as a conservation metric. (H) RNA sequencing by cleavage was the earliest nucleic acid sequencing technology, but RNA sequencing was expensive and difficult until the reverse transcriptase enzyme was commercially available. This allowed RNA to be converted into ‘complementary DNA’ and sequenced by any DNA sequencing technology. HTS allowed the sequencing of all RNA transcripts in a sample and recent methods allow sequencing of whole, single RNA molecules. (I) Methylomics was enabled by bisulfite treatment of DNA, after which any HTS DNA sequencing platform can be used to determine CpG methylation levels. Newer long-read sequencing platforms can directly identify methylated cytosines and other modified bases. (J) Genomes of bacteriophages were assembled by manual alignment, the first being Φx174 in 1977. Bacterial and then yeast genomes were first sequenced from Sanger DNA sequencing data in the 1990s. High-throughput implementations of Sanger sequencing were used for the human genome project, initially completed in 2001. The high-throughput sequencing (HTS) era began with 454 ‘pyrosequencing’ technology, which was sold as ‘next generation sequencing’ (NGS). HTS enabled complete genomes of any organism to be produced, with cost declining and quality of assembly improving with each iteration of HTS technologies. (K) Bioinformatics developed partly because of the abundant gene sequence data being deposited in nucleic acid sequence databases. Sequence alignment tools developed into database searching and genome assembly tools. Most chapters in this book involve some bioinformatic analysis, but those completely dependent on it are indicated.

WHAT IS ‘ENVIRONMENTAL GENOMICS’?

Genomes are connected to the environment of their host organism. The connections can be simple and direct, for example in species that can absorb metabolites directly through their cell wall. A classic example of this from formative research on genetics is the regulation of lactose metabolism by Escherichia coli bacteria. The Lac operon of the E. coli genome produces beta-galactosidase for digesting lactose at a rate determined by environmental levels of lactose (Jacob and Monod 1961). This research could now be called ‘environmental genomics’, although in the 1960s it would have been described as ‘molecular biology’ or ‘functional genetics’. At that time, it was only possible to conduct genetic analysis one gene at a time and the term ‘genomics’ did not exist. Modern environmental genomic techniques allow the entire genomes, messenger RNAs and expressed proteins of whole bacterial communities to be studied in comprehensive detail from single environmental samples. With this information we can evaluate both long- and short-term changes to the environment in response to human and non-human actions. In many cases we can do it quickly, in detail and, increasingly, in the field without requiring a laboratory.

Environmental genomics is the use of information encoded in DNA or RNA to characterise and understand pattern and process in biological systems at a small or large scale. It encompasses studies of microbes, fungi, plants and animals. In the case of applied environmental genomics, the focus is on capture and provision of information relevant to natural resource management. This very broad definition spans sectors from water quality to wild harvest, biodiversity assessment and biosecurity monitoring. It includes such varied specific uses as bear counts in landscapes, risk-assessment based on adaptive potential, synthetic biology for pest control, rapid biodiversity surveys, and more. These diverse applications are united because they take advantage of the rich information in genomes and the technological advances that provide rapid and detailed access to it.

Environmental genomics overlaps with research in related fields, partly because of shared scientific aims or methods, but also because terminology changes with time. For example, the fields of population genetics and molecular ecology were established before genomics methods were prevalent but could now be considered environmental genomics. As a rule of thumb, any environmental research that depends on high-throughput DNA sequencing (HTS), and the associated high-performance computing (HPC), likely falls under the umbrella of environmental genomics. Bioinformatic analysis is part of the workflow, and one hallmark of environmental genomics is that research in this field requires expertise in ecology, genomics, statistical analysis and bioinformatics.

WHAT IS IN THIS BOOK?

Our purpose is to introduce and synthesise the latest applications of genomic technologies for environmental science and management. We aim to emphasise the breadth of applications, highlighting real-world demonstrations, and to provide a one-stop-shop for advanced undergraduate students, postgraduate students and professionals.

We have focused on new applications of genomic technologies to environmental problems. The precise techniques used to study these problems will change with time, but we consider the chosen applications to have real value in solving current environmental issues. This book cannot provide comprehensive coverage of all aspects of environmental genomics. We chose not to cover several areas because they are so well covered by other publications. For example, molecular microbiology is a field of environmental genomics, but is thoroughly served by journals such as Applied Environmental Ecology and comprehensive, classic textbooks (e.g. Madigan et al. 2018). There are likewise new methods that are relevant to environmental genomics that are very well described in existing literature, for example spatially resolved transcriptomics (Larsson et al. 2021) and current genome assembly methods that recover haplotype phases (Kronenberg et al. 2021).

Each chapter of this book describes the application of genomic technologies to environmental problems or ecological research questions. We have asked experts to write chapters that are innovative, not well covered in other textbooks, or that have had significant interest from environmental managers. Some of these environmental managers have written perspectives on how the genomic technologies have enabled them to undertake new or improved work. They are typically non-scientists whose daily work involves use of genomic-derived data to make decisions about the management of natural resources (Chapters 3, 9, 12, 15, 17, 20 and 27). This serves multiple aims. First, it illustrates that indeed genomics already has many and varied, very practical uses. Second, it points to the challenges sometimes faced in taking novel science out of the laboratory and into practical use. Finally, it emphasises the central importance of this translational step, since solving complex environmental problems requires much more than excellent science and innovation: it also requires commitment to partnering with the non-scientific world.

For convenience, we have organised the 27 chapters of this book into six themed sections, with each section framed around an area of application. Section A ‘Biodiversity’ focuses on characterisation of complex biological communities, say for biodiversity or biosecurity monitoring, and largely through DNA metabarcoding technologies, including the hot topic of environmental DNA ‘eDNA’. The uptake of DNA metabarcoding by researchers, government and industry has been rapid, and this section emphasises the important subjects of sampling and survey design (Chapter 1), metrics for measurement of biodiversity in metabarcoding studies (Chapter 2), how to use metabarcoding to characterise trophic interactions among species through the analysis of diet (Chapter 4), and selection and use of metabarcoding assays appropriate to each biological question (Chapter 5).

Section B ‘Life history and population biology’ focuses on development and use of biomarkers for ecological properties of individuals or populations. This includes the key attribute of abundance, which can be estimated through non-invasive sampling (Chapter 10) or analysis of kinship (Chapter 11). Demonstrating the versatility of genomics, those same samples can yield even more detailed insights, including an organism’s age (Chapter 7), lifespan (Chapter 6) and sex (Chapter 8). Finally, Chapter 13 synthesises the fundamental links between evolutionary and ecological processes and their underpinning of population viability.

Section C ‘Adaptation and change’ expands the temporal frame to view current environments and species in historical and future contexts. This recognises that characterisation of evolutionary processes and past environmental changes can help predict future trajectories, evaluate extinction risk, and guide decisions. Chapter 14 evaluates best practice for understanding the capacities of organisms to respond to environmental change through adaptation. Chapter 16 describes ancient DNA and ‘museomics’ applications to answer questions about extinct and extant organisms.

Section D ‘Environmental molecular physiology’ outlines the use of biomarkers for tracking the state and change of organisms or whole communities. Genomic methods provide the ability to rapidly characterise relative levels of nucleic acids with diverse functions. Analysis of the patterns of nucleic acids associated with physiological or community state provides a rapid means of identifying specific individual or community stress. Examples include evaluation of broad-scale epigenetic changes that relate to organismal physiological condition (Chapter 18), and at the other end of the scale, how the structure and metabolic behaviour of whole microbial communities can indicate broader ecosystem condition (Chapter 19).

Section E ‘Spatial genomics’ explores the use of genomics to track the movements of organisms from fine scale, short-term and individual level to the long-term species and population level. This includes the complex interactions between plants and the animals that pollinate them (Chapter 21), and landscape-scale movements of organisms and how they respond to features (Chapter 22). At a longer temporal scale, phylogenomic analysis of whole communities provides an objective means of comparing the evolutionary distinctiveness of communities, which is a useful measure for prioritising landscapes or communities for protection (Chapter 23).

Finally, Section F ‘Biosecurity and disease monitoring’ summarises leading directions in the use of genomics to track invasive species, including pathogens, and to manage them. This theme is not only of great environmental and economic significance, but also highly topical in this era of COVID-19 and other emerging zoonoses. Chapter 24 illustrates the versatility of genomics to support management of the four stages of invasion threat – prevention, eradication, containment and asset protection. Chapter 25 considers current genomic approaches for studying pathogenic organisms in their wildlife hosts. Chapter 26 is novel amongst all other chapters in this book because it discusses how genomes can be manipulated to offer levers for the control of invasive species. This is a challenging area of environmental genomics, not only technically, but also because social acceptance of these technologies is not universal.

Many technical terms are used in this book. Those in bold font are defined in the glossary section.

WHO IS THIS BOOK FOR?

Like the authors of this book, we anticipate that readers are individuals motivated to enact change in environmental science, and to understand and adopt new technologies that will make this possible. Readers will include advanced undergraduate students, postgraduate students, university lecturers and professionals working in natural resource management and policy. For the non-scientist, this text can be read as an introduction to a fast-moving field that undoubtedly is transforming the practice of environmental management with profound relevance to industry, government, and the public. It can also form the basis of a broad and advanced curriculum for university courses. To assist with teaching, authors of each chapter have provided a series of philosophical and practical questions designed to prompt readers to reflect on the chapter content.

Ultimately the goal of this book is to expose a wide audience to the real-world applications of genomics for environmental research. We hope that the discussions stimulated by these 20 focal areas of applied research will break down barriers that prevent successful implementation. It is only with a smooth translational pipeline and excellent communication that we as a community can together face the emerging and accelerating environmental challenges of the coming century.

There is reason for optimism that environmental genomics will catalyse real change in natural resource management, as innovation has consistently driven developments in the field for decades (Fig. 1). As editors it has been an enormous pleasure to work with a group of authors so committed to exposing this exciting and important field to a wider audience. We wish to express our sincere gratitude to them for their efforts, and to acknowledge the non-trivial time and effort they have devoted to this project when they have many competing responsibilities. We hope the result, which is much more than the sum of its parts, makes it worthwhile. Finally, to you the reader, we hope this text provides you with knowledge and confidence to use and extend the power of genomics for understanding and managing our environment.

Olly, Clare, Simon, August 2022

REFERENCES

Almond RE, Grooten M, Peterson T (2020) Living Planet Report 2020-Bending the Curve of Biodiversity Loss. World Wildlife Fund, Gland, Switzerland.

Bak RO, Gomez-Ospina N, Porteus MH (2018) Gene editing on center stage. Trends in Genetics 34(8), 600–611. doi:10.1016/j.tig.2018.05.004

Bar-On YM, Phillips R, Milo R (2018) The biomass distribution on Earth. Proceedings of the National Academy of Sciences 115(25), 6506–6511.

Bessey C, Jarman SN, Simpson T, Miller H, Stewart T, et al. (2021) Passive eDNA collection enhances aquatic biodiversity analysis. Communications Biology 4(1), 1–12. doi:10.1038/s42003-021-01760-8

Higuchi R, Bowman B, Freiberger M, Ryder OA, Wilson AC (1984) DNA sequences from the quagga, an extinct member of the horse family. Nature 312(5991), 282–284. doi:10.1038/312282a0

Höss M, Kohn M, Pääbo S, Knauer F, Schröder W (1992) Excrement analysis by PCR. Nature 359(6392), 199–199. doi:10.1038/359199a0

Jacob F, Monod J (1961) Genetic regulatory mechanisms in the synthesis of proteins. Journal of Molecular Biology 3(3), 318–356. doi:10.1016/S0022-2836(61)80072-7

Kronenberg ZN, Rhie A, Koren S, Concepcion GT, Peluso P, et al. (2021) Extended haplotype-phasing of long-read de novo genome assemblies using Hi-C. Nature Communications 12(1), 1935. doi:10.1038/s41467-020-20536-y

Larsson L, Frisén J, Lundeberg J (2021) Spatially resolved transcriptomics adds a new dimension to genomics. Nature Methods 18(1), 15–18. doi:10.1038/s41592-020-01038-7

Madigan M, Bender K, Buckley D, Sattley W, Stahl D (2018) Brock Biology of Microorganisms. 15th Global Edition. Benjamin Cummins, Boston, USA.

Samuel G (2020) Independent review of the EPBC Act: Interim report. Commonwealth of Australia, Canberra, Australia.

Taberlet P, Coissac E, Pompanon F, Brochmann C, Willerslev E (2012) Towards next-generation biodiversity assessment using DNA metabarcoding. Molecular Ecology 21(8), 2045–2050. doi:10.1111/j.1365-294X.2012.05470.x

Taberlet P, Waits LP, Luikart G (1999) Noninvasive genetic sampling: Look before you leap. Trends in Ecology & Evolution 14(8), 323–327. doi:10.1016/S0169-5347(99)01637-7

Woese CR, Fox GE (1977) Phylogenetic structure of the prokaryotic domain: The primary kingdoms. Proceedings of the National Academy of Sciences 74(11), 5088–5090. doi:10.1073/pnas.74.11.5088

SECTION A

BIODIVERSITY

1Design considerations for eDNA metabarcoding surveys

William Bernard Perry, Kirthana Pillay, Paul George, Georgina Brennan, Abigail Lowe, Laura Jones, Luke Holman, Tom Gibson, Natasha de Vere and Simon Creer

ABSTRACT

There is a great diversity of eDNA metabarcoding studies in the literature and identifying how to design a survey to best suit your needs can be challenging. Design considerations are particularly important given that eDNA metabarcoding can be used to survey biodiversity across a breadth of environments and identifying taxa across the tree of life. Here, we highlight eight burgeoning areas of eDNA metabarcoding research: air, plant-pollinators, soil, diet, microbiome, freshwater, estuarine and marine. We highlight design considerations that are important for specific contexts, while also identifying common denominators across all eDNA metabarcoding surveys. In doing so, we hope to provide both a valuable introduction into eDNA metabarcoding survey design for beginners, gold standards of survey design, and fertile ground for collaboration between research areas which all fall under the umbrella of eDNA metabarcoding.

INTRODUCTION

At some point in our scientific lives, we may be fortunate enough to design an environmental DNA (eDNA) metabarcoding experiment. We refer to metabarcoding sensu (Taberlet, Coissac et al. 2012) and recommend readers review eDNA definitions and references therein according to Bohmann et al. (2014), Creer et al. (2016), Deiner et al. (2017). It can be an overwhelming challenge, but an intuitive puzzle to solve if you break the problem down into base principles in relation to your research questions. If performed correctly, the benefits will be large and immensely satisfying; alternatively, if some key decisions are not made correctly, the results may never materialise, or may be compromised. The aim of this chapter is to highlight some of the key design considerations to leverage high quality metabarcoding data that be effectively interpreted. Simultaneously, we will provide a range of contemporary, exemplar studies from diverse applications to inspire success for the next generation of biodiversity metabarcoding practitioners.

SAMPLING ENVIRONMENT

Air

While the collection of airborne eDNA is relatively new compared to aquatic eDNA, many organisations have been collecting particles from the air for decades, such as national pollen forecasts (Adams-Groom et al. 2002) or monitoring radioactive fallout (Karlsson et al. 2020; Söderström et al. 2002). More recently researchers have used molecular techniques to explore the airborne biodiversity of plants (Brennan et al. 2019), bacteria (Bowers et al. 2011), fungi (Ovaskainen et al. 2020), insects (Roger et al. 2021) and mammals (Clare et al. 2021; Lynggaard et al. 2021). Furthermore, patterns of airborne biodiversity have been linked with observations from terrestrial environments (Bowers et al. 2011; Brennan et al. 2019), allowing exploration of relationships between airborne communities and human health (Yamamoto et al. 2012; Rowney et al. 2021).

Fig. 1.1. A landscape view of the various environments sampled for eDNA and examples of equipment used in sampling them, including (a) Burkard Hirst design volumetric trap for sampling aerial DNA, (b) honey and pollen from insects for sampling DNA of flowering plants, (c) soil corer for sampling DNA of soil biota, (d) gastrointestinal and faecal samples for examining dietary and microbiome DNA, (e) toothbrush used for collecting biofilms and a bottle for collecting DNA in aqueous freshwater, estuarine and marine environments, and (f) a rosette sampler deployed from a research vessel for collecting DNA at different depths in the pelagic marine environment.

There are important considerations when designing an airborne eDNA survey that will influence biodiversity information recovered, including spatial and temporal elements, as well as the choice of sampling equipment. There is a variety of air sampling devices available, from passive samplers such as car cabin filters (Hurley et al. 2019), to targeted air samplers such as Burkard Hirst design volumetric traps (Kraaijeveld et al. 2015; Fig. 1.1a), Burkard Automatic Multi-Vial Cyclone Samplers (Brennan et al. 2019; Ovaskainen et al. 2020), SASS® 3100 Dry Air Samplers (typically used for microbes; Mbareche et al. 2018) and even handheld aerial samplers (de Weger et al. 2020). However, it is important to note that variations in the volume of air sampled and the material that captures particles will influence interpretation of data and downstream processing. For example, particles sampled into water (e.g. Coriolis Micro air sampler; Roger et al. 2021) require a further filtering step to prepare samples for DNA extraction (West and Kimber 2015).

Location and position of sampling devices (ground-level vs. building-top locations) and length of sampling time are directly relevant and influence our understanding of diversity and composition at local and regional scales, and across seasons. In addition, land-use, climatic and meteorological variables will directly influence the ecology of airborne eDNA (such as production and transport of eDNA; Bowers et al. 2011, 2013; Gandolfi et al. 2015; Brennan et al. 2019; Karlsson et al. 2020) and the importance of these variables in shaping the community composition will in turn be influenced by the same sampling decisions. Furthermore, modelling additional meteorological and climatic data will enable robust predictions on airborne community structure (Kurganskiy et al. 2021).

Plant-pollinators

Plant DNA metabarcoding can be used to gain an understanding of plant biodiversity which was not previously possible using morphological identification. It improves our ability to monitor biodiversity (Sjögren et al. 2017), detect rare species (Pornon et al. 2019), and explain community interactions (Thomsen and Sigsgaard 2019). In investigations of plant-pollinator interactions, metabarcoding has been used to uncover the foraging preferences of insects by using pollen from honey (De Vere et al. 2017), brood cells (Gresty et al. 2018), the bodies of insects (Potter et al. 2019) and pollen baskets (Richardson et al. 2015; Bänsch et al. 2020; Fig. 1.1b). Examples of application include revealing the changes in honeybee foraging over decadal timescales (Jones et al. 2021) and understanding levels of generalisation and specialisation within plant-hoverfly networks (Lucas et al. 2018), with implications for forage availability within the landscape and pollination.

However, it is important to consider the nature of the study system before designing a metabarcoding survey. For instance, honey, when contrasted with pollen on the bodies of individual bees, will represent the foraging effort of multiple individuals over a longer period of time (De Vere et al. 2017). This knowledge should influence decisions on the survey method (e.g. transects, timed observations) and how many individuals need to be collected to answer the study’s aims. The survey methods are key to understanding the sampling universe, and what information on plant-pollinator interactions will be captured. The flight period of the insects and the flowering season of the plants, in addition to the maximum foraging distance, will affect the temporal and spatial scope of the survey. For example, while honeybees have been recorded as foraging up to 10 km in florally depauperate areas (Beekman and Ratnieks 2000), other bees such as Lasioglossum have been recorded foraging up to 1 km (Beil et al. 2008).

Metabarcoding, when compared with traditional pollinator survey methods, has an increased temporal range which can capture foraging information over a longer period of time (Arstingstall et al. 2021). Typically, for DNA analysis, insects are caught using a combination of nets and sterile tubes, with nets changed periodically and sterilised between surveys (Bell et al. 2017; Galliot et al. 2017; Lucas et al. 2018). For comparability, surveys should also account for meeting minimum weather conditions appropriate for the target group when sampling. In the laboratory, the insect bodies are washed in lysis buffer, to remove the pollen, and consideration should be given to whether the insect specimen requires morphological identification, which may affect processing strategies. Chapter 21 provides a comprehensive review of the design and implementation of DNA metabarcoding research on pollinators.

Soil

Advances in the use of eDNA have allowed researchers to rapidly expand our understanding of biodiversity in soils, especially unknown components of soil biodiversity and their response to change, with metabarcoding surveys having been conducted at national (Terrat et al. 2017; George et al. 2019) and continental scales (Tedersoo et al. 2014; Delgado-Baquerizo et al. 2018; Delgado-Baquerizo et al. 2018).

Soils are composed of a network of solid aggregates and gas- or water-filled habitable pore space (Ruamps et al. 2011; Totsche et al. 2018). Soil aggregates are formed by the adhesion of clay and soil organic matter (SOM), which progressively forms larger masses with the inclusion of more nutrients, material, and organic compounds over time (Amézketa 1999). SOM ranges from simple sugars to complex carbon structures like lignin (Romero-Olivares et al. 2017; Lehmann et al. 2020) and can strongly influence soil physical structure and microbial communities (Dungait et al. 2018). Some components of soil can interfere with DNA extraction, like humic acids that can bind with DNA (Sagova-Mareckova et al. 2008). To reduce inhibition, soil can be mixed with calcium carbonate (Sagova-Mareckova et al. 2008) or phosphate buffer solutions (Taberlet, Prud’Homme et al. 2012) prior to extraction to neutralise these acids.

For eDNA analyses, soil cores are taken from the organic-rich upper horizons (commonly the first 25 cm) using a soil sampler (Fig. 1.1c), as this is where the majority of microbes reside (Fierer et al. 2003). Both contemporary and older signals from soil animals (e.g. tardigrades, nematodes, annelids, and arthropods) can also be detected (George et al. 2019).

The distribution of aggregates, pore space, and nutrients creates diverse microhabitats within soil, and so very different local communities can be present in a small area. Therefore, it is critical that soil samples, or sub-samples, are aggregated and homogenised, to create a sample representative of the area in question (Taberlet et al. 2012). Considerations must also be made for the collection of metadata. pH (Tedersoo et al. 2014; Delgado-Baquerizo et al. 2018), carbon-to-nitrogen ratio (Griffiths et al. 2011; Tedersoo et al. 2014), soil bulk density, organic matter and macronutrients (e.g. phosphorus, sulphur, and nitrogen) can influence microbial communities. Soil architecture, including measures of porosity and available pore space can also be important, with Carson et al. (2010) highlighting that bacterial diversity increases with decreasing pore connectivity. However, these metrics are often much more difficult to analyse and require more specialised expertise.

Diet

Metabarcoding of diet is a cost-effective alternative to morphological identification of prey items (Elbrecht and Leese 2017, chapter 4) and commonly used when direct observation of feeding behaviours is not possible, especially in aquatic organisms (Sousa et al. 2019). Diet analysis provides insights into ecosystem functioning (Duffy et al. 2007), resource usage (Cristóbal-Azkarate and Arroyo-Rodríguez 2007), anthropogenic dietary change (Sousa et al. 2019) and species interactions (Ingala et al. 2021), all of which can be used for wildlife management (Kowalczyk et al. 2011) and conservation (De Barba et al. 2014).

Faecal material and stomach contents are widely used for dietary metabarcoding (Fig. 1.1d). Collection of faecal material is a non-invasive approach (Symondson 2002; Ando et al. 2020), but DNA quality can be poor compared to stomach contents, due to degradation from digestion. Contamination of faecal samples due to decomposition (Hawlitschek et al. 2018) is also common as samples are exposed to the environment, and can vary with humidity and temperature (Oehm et al. 2011). Decomposers such as bacteria, fungi and arthropods can be hard to discern from true diet content (McInnes et al. 2017), but can be removed bioinformatically if the system is well understood, or minimised by collecting fresh scat (Ando et al. 2018). Careful primer choice, as discussed in Chapter 5, and predator specific blocking primers (Vestheim and Jarman 2008), can also be used to prevent amplifying non-target species (Hawlitschek et al. 2018).

Stomach contents yield higher quality DNA as they are not completely degraded through digestion (Hawlitschek et al. 2018). Obtaining stomach contents by dissecting out the entire gastrointestinal tract is an invasive collection method that requires euthanising the individual. Stomach flushing is an alternative that does not involve euthanising and is more frequently used than forcefully expelling faeces. However, these invasive methods are not recommended when working with large or rare organisms.

Since diet is highly subjected to environmental variability, factors such as seasonality and changes in resource availability must be accounted for. Biological factors such as size and sex of individuals also affect dietary preferences (Lee et al. 2021). Finally, other diet analysis methods such as stable isotope analysis, which characterises long-term diet information, can be used to complement metabarcoding, which only provides a snapshot of current diet (Pompanon et al. 2012).

Microbiome

The gut microbiome has been at the centre of host-microbiome research due the volume and diversity of bacteria, and its functional role in digestion, metabolism and immune response (Cresci and Bawden 2015, chapter 2), with applications ranging from medicine (Cammarota et al. 2020) to aquaculture (Perry et al. 2020). It will therefore be the focus here; however, there is a variety of other host-associated microbiomes to explore.

Even within vertebrates, the gut microbiome shows considerable variation due to factors such as ecology and phylogeny (Colston and Jackson 2016). Variation can also be seen within host species, and within an individual gastrointestinal tract, both in time (Frazier and Chang 2020) and space (Chew et al. 2018). When designing a microbiome study, it is important to establish where samples should be taken (foregut, midgut or hindgut), and how often (hours, weeks, months). These questions will differ wildly between systems, but time series data are often neglected, and can be achieved by using non-invasive sampling of the same individual, or destructive sampling of clonal of familial siblings.

Collection of host metadata is also an important consideration. Factors such as age or life stage (Lim et al. 2019), sex (Sylvia et al. 2017) and social rank (Singh et al. 2019) can all impact host-environment interactions, host physiology and thus host microbiomes. The host’s environment can also impact its microbiome (Kivistik et al. 2020), although it is often hard to tease apart interactions between environment, microbiome and host physiology. Controlling for environmental variables within an experiment can therefore be a powerful tool for focusing on host–microbe interactions. This can be achieved through translocation or common garden experimentation (Uren Webster et al. 2020) and laboratory studies where environmental parameters are tightly controlled (McCoy et al. 2017).

How you take microbiome samples is also an important consideration. Like diet analyses, extracting microbial DNA from a host can be prone to co-amplification of host organellar sequences (Fitzpatrick et al. 2018), however, unlike diet studies, differences in methylation between eukaryotes and prokaryotes can allow for separation (Feehery et al. 2013). Taking samples which minimise host tissue contamination, yet still sample intracellular and closely associated gut bacteria, is, however, the most cost-effective measure.

Finally, microbiome studies often focus on bacteria; however, there have been calls to assess a broader variety of microbes, including microbial eukaryotes (Laforest-Lapointe and Arrieta 2018) and viruses (Shkoporov et al. 2019), which also play important ecological roles.

Freshwater

Freshwater ecosystems are ecologically and topographically diverse, but freshwater population declines continue to outpace those in terrestrial and marine habitats (Reid et al. 2019), making effective eDNA metabarcoding surveys a vital tool for biomonitoring.

When designing a freshwater metabarcoding survey, it is first important to know if the system is lentic or lotic. Effective characterisation of eDNA in each of these habitat types may, for example, require different volumes of water to be filtered or sampling to take place at different times (Bedwell and Goldberg 2020). Sampling protocols should be based on how eDNA behaves in lentic and lotic systems, largely due to differences in transport and degradation.

In lotic systems, eDNA is transported downstream, sometimes for kilometres (Pont et al. 2018; Wacker et al. 2019), and therefore riverine dendritic networks are an important consideration. Transport of eDNA from other sources draining into a lentic system may also affect your results, for example lakes, an aquaculture facility pumping discharge into a local stream, or marine species contamination from wetland birds. However, eDNA transport will depend on a number of factors, including the source taxa, water parameters in the river as well as flow rates (Deiner and Altermatt 2014). To better understand transportation in your study system, it is worth including a positive control as part of your experimental design, which may consist of a species eDNA not found in the study system. Understanding eDNA transport can then inform sampling intervals along a river (e.g. every 100 m, 2 km, or just at the river’s mouth), as identifying taxa location may be compromised at a finer scale due to eDNA transport.

Understanding eDNA transport, or dispersion, in lentic systems is complex, as it does not follow a linear direction as in rivers and streams. This is especially true in larger bodies of water and simply sampling from the shore may not be sufficient (Zhang et al. 2020). Vertebrate cage experiments have demonstrated eDNA detection declines sharply at a distance of 5–10 m (Dunker et al. 2016; Brys et al. 2021). eDNA may also become vertically stratified in larger water bodies, depending on abiotic factors linked with season, and so sampling regimes should examine a multitude of depths (Littlefair et al. 2021). In contrast, an issue with smaller water bodies, or the shallows of larger lakes, is that there may be horizontal barriers preventing the movement of eDNA, such as large mats of vegetation (Biggs et al. 2015). In these situations it is recommended that water is taken underneath or around potential eDNA dispersion barriers (Harper et al. 2018).

What eDNA sample you take is vital, both in lentic and lotic systems, as eDNA in the water, sediment, or even biofilms, are likely to reflect community composition in different ways (Sakata et al. 2020). Picking which of these sampling methods to use will depend on the questions you want to answer, and the taxa you want to identify. Finally, collecting metadata on the system alongside your sample will also be beneficial in understanding persistence of eDNA; for example, pH, UV and temperature can all be important explanatory variables (Strickler et al. 2015; Seymour et al. 2018).

Estuarine

Estuaries are highly dynamic environments which feature spatial and temporal gradients in their physical and chemical conditions, which in turn influence biota (McLusky 1993; McLusky and Elliott 2004). Sampling should therefore aim to capture, or account for, this spatio-temporal variability. The timing of sampling will influence the detection of species due to the changing contribution of eDNA transported from adjacent marine and freshwaters (Fig. 1.1e).

The relative importance of marine and freshwater influence varies with river flow, the state of the tide during each flood to ebb cycle and variation in tidal range during the spring to neap cycle (McLusky and Elliott 2004). There is also evidence that tidal state can influence eDNA assessments of biodiversity in a well-mixed meso/macrotidal estuary (Schwentner et al. 2021). Comparably, the effects of tidal flow have been shown to be minimal in a glacial fjord (Kelly et al. 2018). Sampling should therefore be standardised to specific points in the tidal cycle (Burgoa Cardás et al. 2020). The effects of river flow, like tide, are poorly studied in estuaries, but they can influence diversity, partially due to increased eDNA transport and dilution during high flow events (Milhau et al. 2021; Sales et al. 2021). Seasons with generally lower river flows could be favourable.

The spatial location of samples must also be considered. Despite the potential for eDNA transport homogenising community composition over space, distinct fish communities have been detected at the scale of hundreds of kilometres within an estuary (García-Machado et al. 2021). Therefore, estuarine sample sites should be positioned longitudinally, for example in the lower, middle and upper estuary. Depending on hydrography (McLusky and Elliott 2004), vertical transects may be necessary. Evidence from a fjord shows spatially specific eDNA signals can occur with depth across a halocline, in as little as 4 m (Jeunen et al. 2020). Finally, it is critical to contextualise the results by collecting environmental metadata, such as salinity, turbidity, dissolved oxygen and temperature (Hemingway and Elliot 2002).

Two technical issues present themselves when processing estuarine water samples. First, filters can rapidly clog as estuarine waters are often highly turbid. Many estuaries show a turbidity maxima in their middle and upper reaches, which can vary with tidal cycles (McLusky and Elliott 2004). Therefore, maintaining consistent sample volumes, which can influence species detection (Sigsgaard et al. 2017), may not be possible (Ahn et al. 2020). This can be remedied by using large pore sizes (Robson et al. 2016; Simpfendorfer et al. 2016), pre-filtration of samples (Robson et al. 2016; Stoeckle et al. 2017) and filtration of small water volumes (Kelly et al. 2018; Schwentner et al. 2021). Second, as with soils, estuarine waters feature high concentrations of humic compounds and potentially high levels of chemical pollutants which may inhibit PCR.

Marine

From the rocky shore to the deep sea, marine ecosystems are environmentally heterogenous and biologically diverse, and research has shown that many different types of eDNA sample are required to fully capture biodiversity (Holman et al. 2019; Antich et al. 2020). For example, Koziol et al. (2019) demonstrated that water, sediment, settlement plate and plankton trawl samples showed dramatic differences in their recovered community diversity and sensitivity to certain taxa.

Marine eDNA samples contain biodiversity information relevant to many scales of ecological investigation, demonstrating delineation of biological communities at 10–100s of metres (O’Donnell et al. 2017; West et al. 2020), between habitat types separated by kilometres (Jeunen et al. 2019) and across thousands of kilometres at biogeographic scales (Holman et al. 2021; West et al. 2021). Marine communities also change along depth gradients and studies have shown that eDNA can track and recover marine community changes across depth (Jeunen et al. 2020; Canals et al. 2021). Methods to collect marine eDNA vary in complexity depending on the community or species targeted. For example, it is simple to collect surface water from a pontoon using a plastic bottle or to collect sediment using a handheld corer or grab (Holman et al. 2019). In contrast, collecting eDNA from a remote deep-sea environment requires large expensive ocean-going vessels equipped with oceanographic deep-sea sampling instruments, such as rosettes or gravity corers.

Surveys using eDNA have previously captured seasonal changes in marine biodiversity (Rey et al. 2020; Stoeckle et al. 2021), demonstrating the importance of incorporating sampling across time in seasonal seas. Repeated sampling of eDNA can capture temporal changes in diversity, as eDNA samples represent a ‘snap-shot’ in time of the local biodiversity (Holman et al. 2021). A new approach is to use eDNA samplers that aggregate eDNA across time into a single sample, resulting in a time-integrated eDNA sample. Passive samplers can achieve this using existing eDNA diffusion in marine environments (Kirtane et al. 2020; Bessey et al. 2021), while natural samplers (such as sponges or scavenging organisms) have been shown to be enriched with eDNA from local biota (Mariani et al. 2019; Siegenthaler et al. 2019; Turon et al. 2020). However, it is still unclear what spatio-temporal scale eDNA collected from these sources represents and further work is required to understand what proportion of biodiversity these eDNA sources represent.

It is important to conduct pilot studies or examine the literature to ensure the type of marine eDNA sample taken has good sensitivity for the target taxa of interest. In addition, sampling schemes should account for the nested nature of eDNA samples to ensure inference at larger scales incorporates variance at lower sampling scales. For example, when sampling across a coastline, ensuring that all types of habitat, such as the intertidal and the neritic zone, are represented when the goal is to understand the biodiversity of the whole marine ecosystem. Finally, designs should incorporate temporal replicates when the study ecosystem is affected by seasonality or disturbance, for example in temperate seas or the intertidal.

IMPORTANT CONSIDERATIONS

We often get asked, ‘how many samples/replicates should I analyse?’. In an ideal world, each ecological sample would feature multiple technical replicates (minimum of three subsamples extracted from a sample, with three PCR replicates per subsample, all of which are then sequenced) to achieve accurate representation of a complex DNA sample, as well as multiple ecological replicates, to represent ecological variation. A good summary of replicate types in metabarcoding studies is highlighted by Beentjes et al. (2019) in their fig. 3. Ultimately, the number of replicates will depend on your hypotheses and will reflect the optimal level of work according to resource availability. To make the process easier, start with a simple multiplier table of factors such as sites, replicates, time points, types of sample, and number of markers. Care should also be taken in the design and statistical analysis of studies to avoid pseudo-replication of the ecological samples (Hurlbert 1984). Similarly, how many samples/markers should we multiplex on a single sequencing run, or at what depth should we sequence our eDNA samples? Assessing similar studies, or performing trial runs with lower throughput sequencing chemistries (e.g. Illumina MiSeq Nano) to test taxon accumulation asymptotes versus sequencing effort, are great routes to follow for efficient data collection. Importantly, if you predict comparatively low diversity, multiplexing more samples together with 10% PhiX will likely yield better results due to reduced risk of overclustering on Illumina platforms.

Technical considerations are also extremely valuable, such as negative controls taken at multiple stages of the practical work, for example during field collection, DNA extraction and PCR, which are sequenced along with the samples. Good negative controls and controls of taxa that do not feature in your study system will not only allow you to identify contamination or index hopping, but they can also allow you to bioinformatically filter out spurious taxa. Similarly, mock communities provide a valuable insight into how your metabarcoding pipeline may be skewing the taxa you are detecting. For example, in microbiome studies, a DNA extraction method may favour gram-negative bacteria (Ketchum et al. 2018), which would be detected by a good mock community. Finally, there has also been interest in synthetic amplicon spike-in controls, which are added to samples, and can help with the problem of tag jumping, sample mix-ups and cross-contamination (Tourlousse et al. 2018). With the knowledge these technical considerations provide, you are then able to interpret your results in a more informed manner, both statistically and conceptually.

CONCLUSIONS

Despite the breadth of sampling environments (Fig. 1.1) there are core principles of good metabarcoding survey design which minimise noise and maximise detection of ecological signals. In technical terms, this can be achieved by increasing the number of technical replicates or providing negative and mock community controls to later filter your metabarcoding dataset. However, a large proportion of noise within your dataset could come from sampling strategy. In which case, it is important to minimise sampling noise by homogenising sampling campaigns according to your study system. Additionally, you can collect suites of metadata which can help statistically account for noise and confounding variables. Finally, two of the most common sources of sampling variation are time and space, and so it is important to establish adequate spatio-temporal sampling to leverage representative and powerful eDNA biodiversity data for accurate ecological synthesis.

DISCUSSION TOPICS

1. Imagine you are sampling an entirely new ecosystem. Examples might be the deep lithosphere from a mine shaft, or the atmosphere of a moon of Jupiter. What materials would you use to collect any eDNA, and how would you avoid contamination from DNA found in our familiar environment.

2. Environmental DNA metabarcoding is subject to many sources of sampling contamination. How would you design field sampling protocols that help to identify false positive eDNA identifications?

3. How would you compare DNA metabarcoding results for the same biological question, but with samples taken from different field material?

4. How can you be confident that a negative result for a species in an environmental sample means that the eDNA from that species is not really present?

REFERENCES

Adams-Groom B, Emberlin J, Corden J, Millington W, Mullins J (2002) Predicting the start of the birch pollen season at London, Derby and Cardiff, United Kingdom, using a multiple regression model, based on data from 1987 to 1997. Aerobiologia 18, 117–123.

Ahn H, Kume M, Terashima Y, Ye F, Kameyama S, et al. (2020) Evaluation of fish biodiversity in estuaries using environmental DNA metabarcoding. PLoS ONE 15(10), e0231127. doi:10.1371/journal.pone.0231127

Amézketa E (1999) Soil aggregate stability: A review. Journal of Sustainable Agriculture 14(2–3), 83–151. doi:10.1300/J064v14n02_08

Ando H, Fujii C, Kawanabe M, Ao Y, Inoue T, et al. (2018) Evaluation of plant contamination in metabarcoding diet analysis of a herbivore. Scientific Reports 8(1). doi:10.1038/s41598-018-32845-w

Ando H, Mukai H, Komura T, Dewi T, Ando M, et al. (2020) Methodological trends and perspectives of animal dietary studies by noninvasive fecal DNA metabarcoding. Environmental DNA 2(4), 391–406. doi:10.1002/edn3.117

Antich A, Palacín C, Cebrian E, Golo R, Wangensteen OS, et al. (2020) Marine biomonitoring with eDNA: Can metabarcoding of water samples cut it as a tool for surveying benthic communities? Molecular Ecology 30, 3175–3188. doi:10.1111/mec.15641

Arstingstall KA, DeBano SJ, Li X, Wooster DE, Rowland MM, et al. (2021) Capabilities and limitations of using DNA metabarcoding to study plant–pollinator interactions. Molecular Ecology 30, 5266–5297. doi:10.1111/MEC.16112

Bänsch S, Tscharntke T, Wünschiers R, Netter L, Brenig B, et al. (2020) Using ITS2 metabarcoding and microscopy to analyse shifts in pollen diets of honey bees and bumble bees along a mass-flowering crop gradient. Molecular Ecology 29, 5003–5018. doi:10.1111/mec.15675

De Barba M, Miquel C, Boyer F, Mercier C, Rioux D, et al. (2014) DNA metabarcoding multiplexing and validation of data accuracy for diet assessment: Application to omnivorous diet. Molecular Ecology Resources 14(2), 306–323. doi:10.1111/1755-0998.12188

Bedwell ME, Goldberg CS (2020) Spatial and temporal patterns of environmental DNA detection to inform sampling protocols in lentic and lotic systems. Ecology and Evolution 10(3), 1602–1612. doi:10.1002/ECE3.6014

Beekman M. Ratnieks FLW (2000) Long-range foraging by the honey-bee, Apis mellifera L. Functional Ecology 14(4), 490–496. doi:10.1046/J.1365-2435.2000.00443.X

Beentjes KK, Speksnijder AG, Schilthuizen M, Hoogeveen M, van der Hoorn BB (2019) The effects of spatial and temporal replicate sampling on eDNA metabarcoding. PeerJ 7(7), e7335. doi:10.7717/PEERJ.7335

Beil M, Horn H, Schwabe A (2008) Analysis of pollen loads in a wild bee community (Hymenoptera: Apidae) – a method for elucidating habitat use and foraging distances. Apidologie 39(4), 456–467. doi:10.1051/APIDO:2008021

Bell KL, Fowler J, Burgess KS, Dobbs EK, Gruenewald D, et al. (2017) Applying pollen DNA metabarcoding to the study of plant–pollinator interactions. Applications in Plant Sciences 5(6), 1600124. doi:10.3732/apps.1600124

Bessey C, Jarman SN, Simpson T, Miller H, Stewart T, et al. (2021) Passive eDNA collection enhances aquatic biodiversity analysis. Communications Biology 4(1), 1–12. doi:10.1038/s42003-021-01760-8

Biggs J, Ewald N, Valentini A, Gaboriaud C, Dejean T, et al. (2015) Using eDNA to develop a national citizen science-based monitoring programme for the great crested newt (Triturus cristatus). Biological Conservation 183, 19–28. doi:10.1016/J.BIOCON.2014.11.029

Bohmann K, Evans A, Gilbert MT, Carvalho GR, Creer S, et al. (2014) Environmental DNA for wildlife biology and biodiversity monitoring. Trends in Ecology and Evolution 29(6), 358–367. doi:10.1016/j.tree.2014.04.003

Bowers RM, McLetchie S, Knight R, Fierer N. (2011) Spatial variability in airborne bacterial communities across land-use types and their relationship to the bacterial communities of potential source environments. ISME Journal 5(4), 601–612. doi:10.1038/ismej.2010.167

Bowers RM, Clements N, Emerson JB, Wiedinmyer C, Hannigan MP, et al. (2013) Seasonal variability in bacterial and fungal diversity of the near-surface atmosphere. Environmental Science and Technology 47(21), 12097–12106. doi:10.1021/es402970s

Brennan GL, Potter C, De Vere N, Griffith GW, Skjøth CA, et al. (2019) Temperate airborne grass pollen defined by spatio-temporal shifts in community composition. Nature Ecology and Evolution 3(5), 750–754. doi:10.1038/s41559-019-0849-7

Brys R, Haegeman A, Halfmaerten D, Neyrinck S, Staelens A, et al. (2021) Monitoring of spatiotemporal occupancy patterns of fish and amphibian species in a lentic aquatic system using environmental DNA. Molecular Ecology 30(13), 3097–3110. doi:10.1111/MEC.15742

Burgoa Cardás J, Deconinck D, Márquez I, Torre PP, Garcia-Vazquez E, et al. (2020) New eDNA based tool applied to the specific detection and monitoring of the endangered European eel. Biological Conservation 250, 108750. doi:10.1016/j.biocon.2020.108750

Cammarota G, Ianiro G, Ahern A, Carbone C, Temko A, et al. (2020) Gut microbiome, big data and machine learning to promote precision medicine for cancer. Nature Reviews Gastroenterology and Hepatology 17(10), 635–648. doi:10.1038/s41575-020-0327-3

Canals O, Mendibil I, Santos M, Irigoien X, Rodríguez-Ezpeleta N (2021) Vertical stratification of environmental DNA in the open ocean captures ecological patterns and behavior of deep-sea fishes. Limnology and Oceanography 6, 339–347. doi:10.1101/2021.02.10.430594

Carson JK, Gonzalez-Quiñones V, Murphy DV, Hinz C, Shaw JA, et al. (2010) Low pore connectivity increases bacterial diversity in soil. Applied and Environmental Microbiology 76(12), 3936–3942. doi:10.1128/AEM.03085-09

Chew YM, Lye S, Md. Salleh M, Yahya A (2018) 16S rRNA metagenomic analysis of the symbiotic community structures of bacteria in foregut, midgut, and hindgut of the wood-feeding termite Bulbitermes sp. Symbiosis 76(2), 187–197. doi:10.1007/s13199-018-0544-5

Clare EL, Economou CK, Faulkes CG, Gilbert JD, Bennett F, et al. (2021) eDNAir: Proof of concept that animal DNA can be collected from air sampling. PeerJ 9, e11030. doi:10.7717/peerj.11030

Colston TJ, Jackson CR (2016) Microbiome evolution along divergent branches of the vertebrate tree of life: what is known and unknown. Molecular Ecology 25, 3776–3800. doi:10.1111/mec.13730

Creer S, Deiner K, Frey S, Porazinska D, Taberlet P, et al. (2016) The ecologist’s field guide to sequence-based identification of biodiversity, Methods in Ecology and Evolution 7, 1008–1018. doi:10.1111/2041-210X.12574

Cresci GA, Bawden E (2015) Gut microbiome: What we do and don’t know. Nutrition in Clinical Practice 30, 734–746. doi:10.1177/0884533615609899

Cristóbal-Azkarate J, Arroyo-Rodríguez V (2007) Diet and activity pattern of howler monkeys (Alouatta palliata) in Los Tuxtlas, Mexico: Effects of habitat fragmentation and implications for conservation. American Journal of Primatology 69(9), 1013–1029. doi:10.1002/ajp.20420

Deiner K, Bik HM, Mächler E, Seymour M, Lacoursière-Roussel A, et al. (2017) Environmental DNA metabarcoding: Transforming how we survey animal and plant communities. Molecular Ecology 26, 5872–5895. doi:10.1111/mec.14350

Deiner K, Altermatt F (2014) Transport distance of invertebrate environmental DNA in a natural river. PLoS ONE 9(2), e88786. doi:10.1371/journal.pone.0088786

Delgado-Baquerizo M, Oliverio AM, Brewer TE, Benavent-González A, Eldridge DJ, et al. (2018) A global atlas of the dominant bacteria found in soil. Science 359(6373), 320–325. doi:10.1126/science.aap9516

Delgado-Baquerizo M, Reith F, Dennis PG, Hamonts K, Powell JR, et al. (2018) Ecological drivers of soil microbial diversity and soil biological networks in the Southern Hemisphere. Ecology 99(3), 583–596. doi:10.1002/ecy.2137

Duffy JE, Cardinale BJ, France KE, McIntyre PB, Thébault E, et al. (2007) The functional role of biodiversity in ecosystems: Incorporating trophic complexity. Ecology Letters 10, 522–538. doi:10.1111/j.1461-0248.2007.01037.x

Dungait JAJ, Berhe AA, Gregory AS, Hopkins DW (2018) Physical protection and mean residence time of soil carbon. In Soil and Climate. (Eds R Lal, BA Stewart) pp. 171–182. CRC Press, Boca Raton, FL, USA. doi:10.1201/b21225-6

Dunker KJ, Sepulveda AJ, Massengill RL, Olsen JB, Russ OL, et al. (2016) Potential of environmental DNA to evaluate Northern Pike (Esox lucius) eradication efforts: An experimental test and case study. PLoS One 11(9), 162277. doi:10.1371/journal.pone.0162277

Elbrecht V, Leese F (2017) Validation and development of COI metabarcoding primers for freshwater macroinvertebrate bioassessment. Frontiers in Environmental Science 5(APR). doi:10.3389/fenvs.2017.00011

Feehery GR, Yigit E, Oyola SO, Langhorst BW, Schmidt VT, et al. (2013) A method for selectively enriching microbial DNA from contaminating vertebrate host DNA. PLoS ONE 8(10), 76096. doi:10.1371/journal.pone.0076096

Fierer N, Schimel JP, Holden PA (2003) Variations in microbial community composition through two soil depth profiles. Soil Biology and Biochemistry 35(1), 167–176. doi:10.1016/S0038-0717(02)00251-1

Fitzpatrick CR, Lu-Irving P, Copeland J, Guttman DS, Wang PW, et al. (2018) Chloroplast sequence variation and the efficacy of peptide nucleic acids for blocking host amplification in plant microbiome studies. Microbiome 6(1), 1–10. doi:10.1186/s40168-018-0534-0

Frazier K, Chang EB (2020) Intersection of the Gut Microbiome and Circadian Rhythms in Metabolism. Trends in Endocrinology and Metabolism 31, 25–36. doi:10.1016/j.tem.2019.08.013

Galliot J-N, Brunel D, Bérard A, Chauveau A, Blanchetête A, et al. (2017) Investigating a flower-insect forager network in a mountain grassland community using pollen DNA barcoding. Journal of Insect Conservation 21, 827–837. doi:10.1007/s10841-017-0022-z

Gandolfi I, Bertolini V, Bestetti G, Ambrosini R, Innocente E, et al. (2015) Spatio-temporal variability of airborne bacterial communities and their correlation with particulate matter chemical composition across two urban areas. Applied Microbiology and Biotechnology 99(11), 4867–4877. doi:10.1007/s00253-014-6348-5

García-Machado E, Laporte M, Normandeau E, Hernández C, Côté G, et al. (2021) Fish community shifts along a strong fluvial environmental gradient revealed by eDNA metabarcoding. Environmental DNA 4, 117–134. doi:10.1002/edn3.221

George PBL, Lallias D, Creer S, Seaton FM, Kenny JG, et al. (2019) Divergent national-scale trends of microbial and animal biodiversity revealed across diverse temperate soil ecosystems. Nature Communications 10(1), 1–11. doi:10.1038/s41467-019-09031-1

Gresty CEA, Clare E, Devey DS, Cowan RS, Csiba L, et al. (2018) Flower preferences and pollen transport networks for cavity-nesting solitary bees: Implications for the design of agri-environment schemes. Ecology and Evolution 8(15), 7574–7587. doi:10.1002/ece3.4234

Griffiths RI, Thomson BC, James P, Bell T, Bailey M, et al. (2011) The bacterial biogeography of British soils. Environmental Microbiology 13(6), 1642–1654. doi:10.1111/j.1462-2920.2011.02480.x

Harper LR, Buxton AS, Rees HC, Bruce K, Brys R, et al. (2018) Prospects and challenges of environmental DNA (eDNA) monitoring in freshwater ponds. Hydrobiologia 826(1), 25–41. doi:10.1007/S10750-018-3750-5

Hawlitschek O , Fernandez-Gonzalez A, Balmori-de la Puente A, Castresana J (2018) A pipeline for metabarcoding and diet analysis from fecal samples developed for a small semi-aquatic mammal. PLoS ONE 13(8). doi:10.1371/journal.pone.0201763

Hemingway KL Elliot M (2002) Field methods. In Fishes in Estuaries. (Eds K.L. Hemingway and M. Elliot) pp. 410–509. Blackwell Science Ltd, Oxford, UK.

Holman LE, de Bruyn M, Creer S, Carvalho G, Robidart J, et al. (2019) Detection of introduced and resident marine species using environmental DNA metabarcoding of sediment and water. Scientific Reports 9(1), 1–10. doi:10.1038/s41598-019-47899-7

Holman LE, De Bruyn M, Creer S, Carvalho G, Robidart J, et al. (2021) Animals, protists and bacteria share marine biogeographic patterns. Nature Ecology and Evolution 5(6), 738–746. doi:10.1038/s41559-021-01439-7

Hurlbert SH (1984) Pseudoreplication and the design of ecological field experiments. Ecological Monographs 54(2), 187–211. doi:10.2307/1942661

Hurley KV, Wharton L, Wheeler MJ, Skjøth CA, Niles C, et al. (2019) Car cabin filters as sampling devices to study bioaerosols using eDNA and microbiological methods. Aerobiologia 35(2), 215–225. doi:10.1007/S10453-018-09554-Y

Ingala MR, Simmons NB, Wultsch C, Krampis K, Provost KL, et al. (2021) Molecular diet analysis of neotropical bats based on fecal DNA metabarcoding. Ecology and Evolution 11(12), 7474–7491. doi:10.1002/ECE3.7579

Jeunen G, Lamare MD, Knapp M, Spencer HG, Taylor HR, et al. (2020) Water stratification in the marine biome restricts vertical environmental DNA (eDNA) signal dispersal. Environmental DNA 2(1), 99–111. doi:10.1002/edn3.49

Jeunen GJJ, Knapp M, Spencer HG, Lamare MD, Taylor HR, et al. (2019) Environmental DNA (eDNA) metabarcoding reveals strong discrimination among diverse marine habitats connected by water movement. Molecular Ecology Resources 19(2), 426–438. doi:10.1111/1755-0998.12982

Jones L, Brennan GL, Lowe A, Creer S, Ford CR, et al. (2021) Shifts in honeybee foraging reveal historical changes in floral resources. Communications Biology 4(1), 1–10. doi:10.1038/s42003-020-01562-4

Karlsson E, Johansson AM, Ahlinder J, Lundkvist MJ, Singh NJ, et al. (2020) Airborne microbial biodiversity and seasonality in Northern and Southern Sweden. PeerJ 2020(1), 8424. doi:10.7717/peerj.8424

Kelly RP, Gallego R, Jacobs-Palme E (2018) The effect of tides on nearshore environmental DNA. PeerJ 2018(3), e4521. doi:10.7717/peerj.4521

Ketchum RN, Smith EG, Vaughan GO, Phippen BL, McParland D, et al. (2018) DNA extraction method plays a significant role