Zooplankton — tiny drifting animals like copepods, krill, and jellyfish — are the engines of ocean food webs. They connect microscopic algae to fish, whales, and seabirds, and play a huge role in moving carbon from the surface ocean to the deep sea. Monitoring these communities is therefore critical for understanding how ocean ecosystems are responding to climate change. Collecting zooplankton is simple and fast, but identifying hundreds of species under a microscope is slow, expensive, and requires rare expertise. DNA metabarcoding offers a powerful alternative: by reading short genetic "barcodes" from a whole community sample at once, we can rapidly identify hundreds to thousands of species at a time. However, metabarcoding involves dozens of steps in the field and in the lab, and each step requires critical researcher input: how samples are preserved, which DNA barcode is used, which sequencing machine is used, and how the data are processed. Choices along this workflow can all influence the results, making it hard to compare studies from different labs or time periods. Some of these steps also remain very labour intensive and logistically challenging, limiting the number of samples a single study can collect and process.

Systematically testing steps in eDNA methodology

In a new paper published in Molecular Ecology Resources, Elizaveta Ershova-Menze, Akvaplan-niva researcher and first author on the study, explains: 'We systematically tested how methodological decisions affect what zooplankton communities we detect and how we interpret them. We tracked the effects of choices all the way from sample preservation and DNA extraction through to sequencing platforms and taxonomic assignments to evaluate how much each step shapes the final picture of biodiversity. Importantly, we also developed and validated a streamlined field- and laboratory protocol that significantly cuts both the cost and time needed to process samples, making large-scale, routine monitoring programs far more feasible than before." One of the keys to this was finding a cost-effective and field-friendly alternative to ethanol – the most common way of preserving zooplankton samples, eliminating the need for large volumes of expensive and hazardous chemicals at sea. Samples were also collected in a way that permits the use of large-scale automated workflows in the lab, reducing manual labor and allowing the simultaneous processing of hundreds of samples at a time. Ershova-Menze continues, "The good news on the broader picture: overall biodiversity patterns and community structure were recovered consistently across all the method combinations we tested, supporting the idea that core ecological signals from DNA metabarcoding are robust and can be compared across studies. However, some methodological choices were more cost- and time-efficient, while maintaining a high data quality."

Towards a more standardized and effective environmental monitoring method

These results have real implications for large-scale ocean monitoring programs, which increasingly rely on DNA tools to track biodiversity across national borders and decades. By identifying where methodological choices matter more vs. less, and by offering a practical, affordable workflow that can be adopted widely, this work helps pave the way for standardized, comparable, and ultimately more reliable ocean biodiversity data. The goal is a future where a net tow collected off Norway and one from the North Pacific can be confidently compared in the same database, giving scientists and policymakers a clearer window into how the global ocean is changing. The work is a collaboration between Akvaplan-niva and the Institute of Marine Research.

Collecting zooplankton with a WP2 net in the North Sea onboard R/V Johan Hjort.

Analysing zooplankton samples under the microscope – a tedious and labour-intensive task.

Extracting DNA using a robotic workstation at the IMR Flødevigen Research Station. This instrument can extract DNA from 384 samples in 2 hours, significantly reducing manual labour.

Link to the scientific article: https://onlinelibrary.wiley.com/doi/10.1111/1755-0998.70149