DNA metabarcoding is a method for identifying many species at once from a single environmental or bulk biological sample. Instead of catching, trapping, or visually counting organisms, researchers collect water, soil, air, feces, or pooled specimens, extract the DNA present, amplify a short standardized gene region, and run it through high-throughput sequencing to generate a list of species detected. The technique has reshaped biodiversity monitoring since the early 2010s, making it possible to survey entire communities of fish, insects, fungi, plants, and microbes from a tube of river water or a scoop of sediment. What makes it powerful, and what limits it, depends heavily on the details of how each step is carried out.
From Sample to Species List
The workflow has a few core stages that stay roughly the same whether you are sampling pond water or bat droppings. First, DNA is extracted from the sample. In “bulk” metabarcoding, this means grinding up a physical collection of organisms, like the insects caught in a trap. In environmental DNA (eDNA) metabarcoding, the DNA comes from traces organisms have shed into their surroundings: skin cells, mucus, feces, pollen, or decaying tissue floating in water or settling on surfaces. Either way, you end up with a soup of genetic material from many species mixed together.
Next comes amplification. Researchers use polymerase chain reaction (PCR) to copy a specific short stretch of DNA millions of times over. This short stretch, called a barcode region, is chosen because it is similar enough across many species that the same set of PCR primers can latch onto it, yet variable enough between species that the sequence differences allow identification. The amplified DNA is then sequenced on a high-throughput platform, producing millions of short reads. Automated liquid-handling robots can process hundreds of samples in parallel, making large-scale monitoring programs feasible at modest cost.1Environmental Science and Ecotechnology. Standardized high-throughput biomonitoring using DNA metabarcoding: Strategies for the adoption of automated liquid handlers
The raw sequence data then passes through bioinformatics pipelines that merge overlapping reads, strip out the primer sequences, filter low-quality data, remove chimeric sequences (artificial hybrids formed during PCR), and cluster the remaining reads into groups that represent distinct taxa. Those groups are compared against reference databases of known barcode sequences to assign species or genus names.2PubMed Central. A pile of pipelines: An overview of the bioinformatics software for metabarcoding data analyses The output is essentially a table: which species showed up in each sample and how many sequence reads belonged to each one.
Why Primer Choice Makes or Breaks a Survey
The short gene region you target determines which organisms you can detect, and no single primer pair catches everything. For animals, the most widely used barcode region is a stretch of the mitochondrial cytochrome c oxidase I (COI) gene. For plants, chloroplast genes like rbcL and matK are common. For fungi, the internal transcribed spacer (ITS) region is standard. For fish specifically, a short fragment of mitochondrial 12S or 16S ribosomal RNA genes often performs better because fish DNA in water samples tends to be degraded into small pieces.
The problem is that primers are not perfectly universal. Highly degenerate primers, designed to bind across a broad range of organisms, can introduce noise by also amplifying non-target DNA like bacteria or algae in samples where you only wanted animals.3Metabarcoding and Metagenomics. A DNA metabarcoding protocol for hyporheic freshwater meiofauna: Evaluating highly degenerate COI primers and replication strategy Conversely, primers designed to be more specific can miss entire groups. In freshwater insect surveys, primer mismatches in the binding region have caused near-complete failure to detect caddisflies (Trichoptera), even when those insects were physically present at sampling sites.4Ecological Indicators. Environmental DNA sequences from aquatic insects indicate freshwater ecosystem state Specialized primers targeted at narrower groups can reveal diversity that universal primers miss entirely. In one study, primers designed specifically for a class of golden-brown microalgae detected 179 distinct genetic variants, whereas previous surveys using broad eukaryotic primers had found the group only rarely.5PubMed Central. Genetic Markers for Metabarcoding of Freshwater Microalgae: Review
This means researchers regularly run multiple primer sets on the same set of samples, targeting different barcode regions to capture different parts of the community. A marine biodiversity study, for instance, might use one assay aimed at fish, another at crustaceans, and a third using a broad COI primer for invertebrates, collectively yielding over a thousand distinct taxonomic units across the three assays.6ICES Journal of Marine Science. Environmental DNA metabarcoding describes biodiversity across marine gradients
Surveying Life in Water
Aquatic ecosystems were the first major arena for eDNA metabarcoding, and they remain the most heavily studied. Organisms in water constantly shed DNA, and because DNA dissolves and disperses in the water column, a filtered water sample can contain genetic traces of dozens of species that live nearby. This is far less disruptive than electrofishing, trawling, or netting, which can injure animals and disturb habitats.
A study of the Danjiang River in China illustrates the potential. Researchers identified 59 fish species across eight orders from eDNA water samples, a richness significantly higher than what traditional cage trapping detected in the same stretch of river. The method also picked up eight rare species and two exotic ones.7Scientific Reports. eDNA metabarcoding reveals differences in fish diversity and community structure in Danjiang River Marine applications have expanded rapidly as well, with eDNA metabarcoding now used to monitor coral reefs, deep-sea habitats, and open-ocean communities in ways that were previously impractical.8PubMed Central. Environmental DNA Metabarcoding in Marine Ecosystems: Global Advances, Methodological Challenges, and Applications in the MENA Region
One persistent caveat with aquatic eDNA is that DNA travels. In rivers, shed DNA drifts downstream before degrading, which means a detection at one site does not necessarily mean the species lives right there. Transport models now account for flow rate, sedimentation, and degradation to estimate how far downstream a detection might be from the actual organism.9PubMed. Predicting downstream transport distance of fish eDNA in lotic environments Another source of false positives in rivers is wastewater discharge. If DNA from aquarium fish or food waste enters a sewer system, it could theoretically show up downstream. However, research on wastewater treatment plants has shown that standard treatment effectively eliminates exogenous DNA, with no detections in effluent or downstream samples after spiking influent with target DNA.10Knowledge and Management of Aquatic Ecosystems. Wastewater treatment effectively eliminates exogenous eDNA, preventing downstream false-positive detections
Figuring Out What Animals Eat
One of the most productive applications of metabarcoding has nothing to do with water sampling. By extracting and sequencing DNA from animal feces, researchers can reconstruct diets with a level of detail that visual identification of half-digested prey items never achieved. The approach works across herbivores, carnivores, and omnivores, and it has become a standard tool in ecology, conservation, and behavioral research.11PubMed Central. Dietary DNA Metabarcoding From Animal Fecal Samples
Dietary metabarcoding is especially valuable for species that are difficult to observe directly. A study on a globally endangered bird used metabarcoding of 60 fecal samples to characterize its diet across three life stages: breeding, post-breeding wandering, and overwintering. The analysis revealed that flies were the dominant animal-derived food, followed by moths and spiders, with plant material also featuring prominently.12PubMed Central. Metabarcoding of fecal DNA reveals the broad and flexible diet of a globally endangered bird That kind of seasonal dietary resolution, distinguishing what a bird eats during breeding versus migration, would be nearly impossible with traditional gut-content analysis, which requires killing the animal.13Environmental DNA. Methodological trends and perspectives of animal dietary studies by noninvasive fecal DNA metabarcoding
Biodiversity from Thin Air
The newest frontier for metabarcoding is literally the atmosphere. Land-dwelling animals, plants, fungi, and bacteria leave DNA traces in the air through shed skin cells, pollen, spores, and other biological particles. Passive air samplers, which are often little more than sticky surfaces or fine filters left exposed for a set period, can collect these traces for sequencing.14Trends in Ecology & Evolution. Transforming terrestrial biodiversity surveys using airborne eDNA
This approach is still young, but early results are striking. Airborne eDNA studies have successfully detected organisms from insects to mammals in terrestrial ecosystems. One study demonstrated that airborne eDNA could detect organisms across the tree of life, track changes in abundance that matched traditional monitoring data, and reveal land-use-driven regional biodiversity decline in a boreal ecosystem over more than three decades.15PubMed Central. Airborne eDNA captures three decades of ecosystem biodiversity Fungal communities in particular are well suited to airborne sampling because they release massive numbers of spores. Recent work has shown that passive air samplers can capture near-real-time variation in plant pollination activity through accumulated pollen DNA, with parallel potential for tracking fungal spore release patterns.16Methods in Ecology and Evolution. Understanding variability and complexity of airborne fungal eDNA Contamination control is a significant challenge, though, since airborne DNA from distant sources or laboratory environments can easily drift into samples.
Freshwater Health Assessment
Governments have long used aquatic insect communities as indicators of water quality. Mayflies, stoneflies, and caddisflies are sensitive to pollution, so their presence or absence tells you something about ecosystem health. Traditionally, trained technicians collect insects with nets, sort them under microscopes, and identify them by morphology. Metabarcoding offers an alternative that can detect far more taxa from the same sites.
In one comparative study, eDNA metabarcoding revealed 63 families and 75 genera of aquatic insects at surveyed sites, roughly double what traditional net sampling found, with the biggest gains in groups like true flies and water bugs that are difficult to identify morphologically. Environmental assessment indices calculated from the eDNA data were more sensitive to pollution gradients than those calculated from net sampling.17PubMed Central. Aquatic insect community structure revealed by eDNA metabarcoding derives indices for environmental assessment The technique is not a perfect replacement, however. As noted earlier, primer mismatches can cause blind spots for certain insect orders, and the overlap between metabarcoding detections and traditional kick-net sampling can be moderate, with one study finding only about half of families detected in common for stoneflies.4Ecological Indicators. Environmental DNA sequences from aquatic insects indicate freshwater ecosystem state The two methods each catch things the other misses, which is why many monitoring programs are exploring combined approaches rather than outright replacement.
Finding Rare, Invasive, and Threatened Species
Metabarcoding’s high sensitivity makes it particularly useful at the extremes of species abundance. Rare and endangered species that are hard to spot in the wild may still leave detectable DNA in their environment. eDNA metabarcoding is now recognized as a non-invasive way to detect such species across aquatic, terrestrial, and atmospheric ecosystems.18PubMed Central. Environmental DNA (eDNA) Technology in Biodiversity and Ecosystem Health Research: Advances and Prospects For conservation, this means surveys can confirm the presence of elusive species in a habitat without the physical disturbance that trapping or netting causes.19Journal for Nature Conservation. Environmental DNA (eDNA): Powerful technique for biodiversity conservation
At the other end of the spectrum, the same water sample that detects a native endangered fish can simultaneously flag an invasive species that has just arrived. Because metabarcoding sequences everything in the sample, it functions as a broad surveillance net. Researchers have proposed integrating invasive species risk assessments directly into metabarcoding reference databases, so that when a sequence matches a known invasive organism, it triggers an alert.20PubMed Central. Integrating invasive species risk assessment into environmental DNA metabarcoding reference libraries Early detection of incipient invasions through routine water sampling is one of the most practically valuable uses of the technology, since eradication becomes exponentially harder once an invasive species establishes a breeding population.
Looking Back Thousands of Years
DNA does not only persist in water and air. It also accumulates in sediment layers over centuries and millennia, preserved in an oxygen-poor environment at the bottom of lakes and oceans. By extracting ancient DNA from dated sediment cores and running metabarcoding, researchers can reconstruct past ecosystems in remarkable detail.
A study of sediment cores from the Yellow Sea used ancient DNA metabarcoding to trace phytoplankton community changes through the mid-to-late Holocene period, linking shifts in species composition to the intensification of a warm ocean current.21PubMed Central. Warm current intensification altered phytoplankton communities in the Yellow Sea: insights from sedimentary ancient DNA metabarcoding On land, sedimentary ancient DNA from a lake in Russia’s Kola Peninsula reconstructed vegetation changes spanning roughly 13,300 years, tracking the progression from post-glacial plant colonization through the Younger Dryas cold snap, into birch forest formation, a mid-Holocene peak in plant diversity with warm-loving species like bird cherry and aspen, and finally the appearance of agricultural weeds signaling human influence.22Journal of Quaternary Science. Vegetation and environmental dynamics in the central part of the Kola Peninsula during the past 13.3 ka as reflected by ancient plant DNA on sediments from Lake Imandra This kind of long-term ecological reconstruction helps scientists understand how ecosystems respond to climate shifts, which informs predictions about ongoing warming.
Wildlife Forensics and Disease Surveillance
Metabarcoding also has a forensic side. In wildlife crime, seized products like traditional medicines, dried meat, or powders often contain heavily processed material from multiple species, making visual identification impossible. Metabarcoding can simultaneously detect DNA from many species in these degraded mixtures, including endangered animals whose trade is illegal.23PubMed Central. Advances in DNA metabarcoding for food and wildlife forensic species identification
In public health, a promising development is the use of metabarcoding for mosquito-borne disease surveillance. Rather than identifying mosquitoes one by one under a microscope and testing each individually for viruses, researchers can process bulk mosquito collections through a multi-locus metabarcoding approach. One such workflow simultaneously identified the mosquito species present, screened for circulating viruses, and identified the vertebrate hosts the mosquitoes had fed on, detecting 18 vertebrate species from their blood meals. This provides a comprehensive picture of disease transmission pathways from a single analytical run.24PubMed Central. An All-in-One Metabarcoding Approach to Mosquito and Arbovirus Xenosurveillance
The Problem with Counting
Metabarcoding excels at answering “what is here?” but struggles with “how much of each?” The number of sequence reads assigned to a species is influenced by many factors beyond how abundant that species actually is: how much DNA it sheds, how well the primers bind to its DNA, how efficiently its DNA amplifies during PCR, and how long its DNA persists in the environment. This means read counts do not straightforwardly translate to biomass or population size.
The relationship between read abundance and actual abundance is real but noisy. Studies comparing eDNA read counts to physical capture data in drained ponds have found positive correlations between the two, but the correlations are sometimes weak and both methods carry independent biases that make discrepancies hard to interpret.25Metabarcoding and Metagenomics. Read counts from environmental DNA (eDNA) metabarcoding reflect fish abundance and biomass in drained ponds Part of this mismatch traces to primer-template mismatches, which have a phylogenetic pattern: closely related species tend to share similar mismatches. For one commonly used barcode gene, COI, incorporating information about the number and type of primer mismatches into statistical models improved biomass estimates, while a different marker, 16S, showed a significant correlation between reads and biomass without needing such correction.26Metabarcoding and Metagenomics. Does phylogeny explain bias in quantitative DNA metabarcoding?
For diet studies, researchers face a related choice: should they report relative read abundance (the proportion of reads assigned to each prey taxon) or frequency of occurrence (how many samples a prey item showed up in)? Occurrence data are often promoted as more conservative, but simulation work suggests that relative read abundance can provide a more accurate picture of population-level diet even when moderate recovery biases exist, though it becomes unreliable when biases affect the most common prey items.27PubMed Central. Counting with DNA in metabarcoding studies: How should we convert sequence reads to dietary data? In practice, many researchers report both and let the reader draw their own conclusions.
Portable Sequencing in the Field
Most metabarcoding to date has relied on large benchtop sequencers, particularly Illumina platforms, which are accurate and produce enormous amounts of data but require well-equipped laboratories. Nanopore sequencing, which uses a device about the size of a USB stick, is changing that equation. The technology reads DNA by pulling single strands through tiny protein pores and measuring electrical current changes, which means it can be used in the field with a laptop.
In a side-by-side comparison of nanopore and Illumina sequencing for eDNA metabarcoding in a biodiverse tropical setting, both platforms detected all four primary terrestrial vertebrate classes. Nanopore actually detected more unique taxa: 18 species were found by both technologies, seven were detected only by nanopore, and just one was detected only by Illumina.28PLoS One. Evaluation of nanopore sequencing for increasing accessibility of eDNA studies in biodiverse countries The higher per-read error rate of nanopore sequencing is a known trade-off, but for many monitoring questions the ability to get results within hours in a remote field station, rather than weeks after shipping samples to a central lab, is worth it. This portability is especially relevant for biodiversity-rich countries in the tropics, where centralized sequencing infrastructure is often limited and the logistics of cold-chain sample transport can be prohibitive.
Reference Database Gaps
A metabarcoding survey can only identify a species if that species already has a barcode sequence deposited in a reference database. For well-studied groups like birds, mammals, and freshwater fish in Europe and North America, coverage is reasonably good. For invertebrates, fungi, tropical taxa, and marine organisms in less-studied regions, the databases have major holes. The Danjiang River fish survey, for instance, noted that local databases need improvement before eDNA metabarcoding can reach its full potential.7Scientific Reports. eDNA metabarcoding reveals differences in fish diversity and community structure in Danjiang River Similarly, studies of freshwater meiofauna have highlighted significant reference database gaps for tiny invertebrate groups.3Metabarcoding and Metagenomics. A DNA metabarcoding protocol for hyporheic freshwater meiofauna: Evaluating highly degenerate COI primers and replication strategy
When a sequence has no close match in the database, it can sometimes be assigned to a genus or family but not a species. In the worst case, it simply shows up as “unknown.” This matters because the species most likely to be missing from databases are often the ones of greatest ecological interest: rare endemics, undescribed species, and organisms in understudied ecosystems. Large-scale barcoding campaigns, which collect specimens, identify them morphologically, and deposit their sequences, remain essential infrastructure for the field. Without them, metabarcoding produces a census with blank name tags.