Identifying Insect Communities: Testing DNA Tests

A fishing spider (Dolomedes tenebrosus) at C&O Canal National Historical Park

NPS

Identifying bugs can be a tricky process that takes lots of time and expertise. And while diverse plant communities and insect communities are strongly linked, we know much less about insect communities than we do plants. That’s why C&O Canal recently sponsored a research project (made possible through an agreement with the Chesapeake Watershed Cooperative Ecosystem Studies Unit) to look at a new way to quickly identify a large number of insects and arachnids mixed together: genetics. Specifically, DNA metabarcoding and high-throughput gene sequencing (see “Terminology” section below).

Manual Bug ID Challenges

If you have an assortment of bugs, manually identifying them based on physical characteristics (called morphological identification) typically involves sorting them into smaller and smaller groups based on body structure. For example, by first dividing things with wings from things without wings. Then into very basic groups like beetles, ants, flies, and wasps. Then further dividing according to shape, color, number of legs, and any other differences. And so on. To get down to a species level identification, high-powered magnification and taxon-specific expertise is often needed. Spiders are particularly challenging to identify unless you have an adult specimen, preferably several. Even then, some taxa are poorly known and taxonomic keys are not yet developed. So, you can see why there might be interest in other ID methods.

DNA versus Manual ID

To learn more about how genetic identification might work for bulk groups of insects and spiders, insects were collected at a subset of CHOH’s Inventory & Monitoring forest vegetation plots and delivered to the Molecular Ecology Research Unit of the Appalachian Laboratory (AL), University of Maryland Center for Environmental Science. Collaborators there sequenced the DNA of the collected bugs and compared the results to a global arthropod reference database representing 87,120 species (from 538,864 trimmed and dereplicated sequences) and made identifications down to the family level.
The result? Genetic analysis detected a total of 184 arthropod families from 38 orders. For spiders, all 16 families detected using manual identification were also detected using genetic analysis. On average, 68.7 percent of sequences were classified to family, with wide variation in the proportion of sequences classified to family per sample, ranging from 17.9 to 93.6 percent.
Analysis is still ongoing, but in this first attempt, genetic identification missed the mark for certain taxa. However, there are many ways the genetic method could be improved, and those improvements could eventually make it the preferred method. Or perhaps an equal partner to manual identification. The genetic method has these significant advantages:

A web-building spider at C&O Canal National Historical Park

NPS

• the use of DNA metabarcoding allows researchers to tackle more comprehensive insect biodiversity studies
• it can also detect species that are often mis-identified manually, especially among similar looking groups and with immature specimens
• it can process hundreds to thousands of samples with minimal cost and effort
• it could greatly improve biological replication among field studies

There’s still some fine tuning to be done though. For example, there was some disagreement between the two methods with respect to which families were detected among specific samples. False negative rates were generally high (but were negatively related to the number of sequences identified from the sample, known as "sequencing depth”) and genetic analysis also identified a taxonomic group not known to be present in the region. Future efforts toward improving the sensitivity and specific genetic markers used are expected to further facilitate the use of these methods. These could include changes to the method and improvements to the infrastructure required for analysis:

• use of multiple metagenetic markers
• greater sequencing depth
• improved sample homogenization methods

• addition of new reference sequences to available public databases
• development of sequence classification methods with increased precision

More Genetics in Our Future

The C&O Canal plans to continue a new phase of this research project to further refine results and hopes to investigate other questions at the park using genetic methods in the future. Elsewhere in the region, NCRN I&M is currently studying how well environmental DNA (eDNA, defined below) identification of macroinvertebrates compares to manual identification in its long-term stream biota monitoring work. Environmental DNA has also been used to look at aquatic macroinvertebrates and fish in Rock Creek Park and similar watersheds at Catoctin and Prince William to determine if it can serve as an indicator of stream health in restored streams. It’s also been used to look for rare or cryptic species like the endangered Hay’s spring amphipod (Stygobromus hayi) at Rock Creek Park, and at Catoctin to detect evidence of the highly invasive freshwater diatom Didymo (Didymosphenia geminata).

The CW CESU promotes stewardship and integrated ecosystem management of natural and cultural resources in the Chesapeake Watershed through collaborative research, technical assistance, and education. To learn more about the CW CESU, please contact Danny Filer at 301-491-2465 or browse the CW CESU Newsletter.

Terminology

• Arthropods, which include insects, spiders, and crustaceans, make up around 80% of all known animal species. They are invertebrate animals having an exoskeleton, a segmented body, and paired jointed appendages.
• DNA metabarcoding and high-throughput sequencing (Liu et al. 2019) is a developing approach to multiple species identification from mixed samples (like environmental DNA/eDNA or bulk DNA) based on high-throughput sequencing (HTS) of a specific DNA/genetic marker. The amount of DNA sequence data derived by HTS allows taxonomy to be rapidly assigned to many species present in a sample.
• eDNA or environmental DNA is DNA extracted from environmental samples (soil, water, air) without isolating a target organism.
• Genetic markers are DNA sequences with known physical locations on chromosomes. They are points of variation that can be used to identify individuals or species, or that are associated with an inherited disease.